.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/experiment_pipeline_lisn.py"
.. LINE NUMBERS ARE GIVEN BELOW.
.. only:: html
.. note::
:class: sphx-glr-download-link-note
:ref:`Go to the end <sphx_glr_download_auto_examples_experiment_pipeline_lisn.py>`
to download the full example code.
.. rst-class:: sphx-glr-example-title
.. _sphx_glr_auto_examples_experiment_pipeline_lisn.py:
Running an Experiment on a dataset using pipeline API
=====================================================
This notebook demonstrates running an experiment to search for hyperparameters and saving the scores of the experiment for a specified set of parameters using `cartodata.pipeline.PipelineExperiment` for `lisn` dataset.
First we will define necessary global variables.
.. GENERATED FROM PYTHON SOURCE LINES 9-26
.. code-block:: Python
from pathlib import Path # noqa
ROOT_DIR = Path.cwd().parent
SOURCE = "authors"
NATURE = "articles"
# The directory where the artifacts of the experiment will be saved
TOP_DIR = ROOT_DIR / "experiment_pipeline_lisn"
# The directory where dataset.yaml files reside
CONF_DIR = ROOT_DIR / "conf"
# The directory where files necessary to load dataset columns reside
INPUT_DIR = ROOT_DIR / "datas"
TOP_DIR
.. rst-class:: sphx-glr-script-out
.. code-block:: none
PosixPath('/builds/2mk6rsew/0/hgozukan/cartolabe-data/experiment_pipeline_lisn')
.. GENERATED FROM PYTHON SOURCE LINES 27-31
Initialize Parameter Iterator
-----------------------------
We will initilialize a parameter iterator to iterate through our parameters. We have two options `GridIterator` and `RandomIterator`.
.. GENERATED FROM PYTHON SOURCE LINES 31-39
.. code-block:: Python
from cartodata.model_selection.iterator import GridIterator, RandomIterator # noqa
help(GridIterator)
""
help(RandomIterator)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Help on class GridIterator in module cartodata.model_selection.iterator:
class GridIterator(ParameterIteratorBase)
| GridIterator(df=None, csv_filepath=None, params_dict=None, index_col=0)
|
| A parameter iterator class that iterates through a given set of
| parameters in order.
|
| Method resolution order:
| GridIterator
| ParameterIteratorBase
| builtins.object
|
| Methods defined here:
|
| __init__(self, df=None, csv_filepath=None, params_dict=None, index_col=0)
| Initializes the params_frame either by the specified ``df``, from the
| specified ``csv_filepath`` or the ``params_dict``.
|
| Parameters
| ----------
| df: pandas.DataFrame
| the dataframe that contains the list of parameter sets.
| csv_filepath: str, Path
| the path of the .csv file that contains the list of parameter sets.
| params_dict: dict
| dictionary of possible parameters specified as key: list pairs.
| index_col: int
| index column if a csv_filepath is specified.
|
| ----------------------------------------------------------------------
| Methods inherited from ParameterIteratorBase:
|
| has_next(self)
| Checks if there is a set of parameters that is that used yet.
|
| Returns
| -------
| bool
| True if there still is a set of parameters that is not used yet;
| False otherwise.
|
| head(self, n=10)
| Views top `n` entries in the `params_frame.`
|
| Parameters
| ----------
| n: int
| the number of rows to view.
|
| next(self)
| Returns the next set of values as dictionary.
|
| Removes ``selected`` column before returning.
|
| Example set of values:
| {
| "id": 7b5e167fb3242dacae5c,
| "robustseed" : 0,
| "dataset" : "lisn",
| "min_df" : 10,
| "max_df" : 0.5,
| "max_words" : "None",
| "vocab_sample" : "None",
| "filter_min_score" : 3,
| "projectionnD" : "lsa",
| "num_dim" : 50,
| "projection2D": "umap",
| "n_neighbors": 15,
| "min_dist": 0.1,
| "metric": "euclidean",
| "clustering" : "kmeans",
| "base_factor" : 3,
| "weight_name_length" : 0,
| }
|
| set_initial_executed(self)
| Check params_frame for the rows that are already run and sets
| `nb_initial_executed` so that these row are not considered for `nbmax`
| value specified for `stop` function.
|
| stop(self, nbmax=None)
| Checks if stopping condition is satisfied. If the experiment is run
| `nbmax` times or there are no other set of parameters to test, stopping
| condition is satisfied.
|
| Parameters
| ----------
| nbmax: int, default=None
| maximum number of experiments to run.
|
| Returns
| -------
| bool
| True if stopping condition is satisfied; False otherwise.
|
| to_csv(self, filepath)
| Persists `params_frame` to the specified `filepath`.
|
| Parameters
| ----------
| filepath: str, Path
| the path to save the contents of params_frame.
|
| transform(self, x)
|
| ----------------------------------------------------------------------
| Readonly properties inherited from ParameterIteratorBase:
|
| columns
| Returns the column names of `params_frame`.
|
| ----------------------------------------------------------------------
| Data descriptors inherited from ParameterIteratorBase:
|
| __dict__
| dictionary for instance variables (if defined)
|
| __weakref__
| list of weak references to the object (if defined)
Help on class RandomIterator in module cartodata.model_selection.iterator:
class RandomIterator(ParameterIteratorBase)
| RandomIterator(df=None, csv_filepath=None, params_dict=None, index_col=0, seed=None)
|
| A parameter iterator class that iterates through a given set of
| parameters by random selections.
|
| Method resolution order:
| RandomIterator
| ParameterIteratorBase
| builtins.object
|
| Methods defined here:
|
| __init__(self, df=None, csv_filepath=None, params_dict=None, index_col=0, seed=None)
| Initializes the params_frame either by the specified ``df``, from the
| specified ``csv_filepath`` or the ``params_dict``.
|
| Parameters
| ----------
| df: pandas.DataFrame
| the dataframe that contains the list of parameter sets.
| csv_filepath: str, Path
| the path of the .csv file that contains the list of parameter sets.
| params_dict: dict
| dictionary of possible parameters specified as key: list pairs.
| index_col: int
| index column if a csv_filepath is specified.
| seed: int
| seed for random generator
|
| ----------------------------------------------------------------------
| Methods inherited from ParameterIteratorBase:
|
| has_next(self)
| Checks if there is a set of parameters that is that used yet.
|
| Returns
| -------
| bool
| True if there still is a set of parameters that is not used yet;
| False otherwise.
|
| head(self, n=10)
| Views top `n` entries in the `params_frame.`
|
| Parameters
| ----------
| n: int
| the number of rows to view.
|
| next(self)
| Returns the next set of values as dictionary.
|
| Removes ``selected`` column before returning.
|
| Example set of values:
| {
| "id": 7b5e167fb3242dacae5c,
| "robustseed" : 0,
| "dataset" : "lisn",
| "min_df" : 10,
| "max_df" : 0.5,
| "max_words" : "None",
| "vocab_sample" : "None",
| "filter_min_score" : 3,
| "projectionnD" : "lsa",
| "num_dim" : 50,
| "projection2D": "umap",
| "n_neighbors": 15,
| "min_dist": 0.1,
| "metric": "euclidean",
| "clustering" : "kmeans",
| "base_factor" : 3,
| "weight_name_length" : 0,
| }
|
| set_initial_executed(self)
| Check params_frame for the rows that are already run and sets
| `nb_initial_executed` so that these row are not considered for `nbmax`
| value specified for `stop` function.
|
| stop(self, nbmax=None)
| Checks if stopping condition is satisfied. If the experiment is run
| `nbmax` times or there are no other set of parameters to test, stopping
| condition is satisfied.
|
| Parameters
| ----------
| nbmax: int, default=None
| maximum number of experiments to run.
|
| Returns
| -------
| bool
| True if stopping condition is satisfied; False otherwise.
|
| to_csv(self, filepath)
| Persists `params_frame` to the specified `filepath`.
|
| Parameters
| ----------
| filepath: str, Path
| the path to save the contents of params_frame.
|
| transform(self, x)
|
| ----------------------------------------------------------------------
| Readonly properties inherited from ParameterIteratorBase:
|
| columns
| Returns the column names of `params_frame`.
|
| ----------------------------------------------------------------------
| Data descriptors inherited from ParameterIteratorBase:
|
| __dict__
| dictionary for instance variables (if defined)
|
| __weakref__
| list of weak references to the object (if defined)
.. GENERATED FROM PYTHON SOURCE LINES 40-41
We define the set of parameters that we want to test. Phase parameters should be specified using phase name and list of dictionaries for phase parameters.
.. GENERATED FROM PYTHON SOURCE LINES 41-73
.. code-block:: Python
from cartodata.phases import PhaseProjectionND, PhaseProjection2D, PhaseClustering
params = {
"robustseed" : [0],
"authors__filter_min_score": [4],
"filter_min_score": [6],
"words__column_names": [["en_keyword_s", "en_domainAllCodeLabel_fs",
"en_abstract_s", "en_title_s"
], ["en_abstract_s", "en_title_s",
"en_keyword_s", "en_domainAllCodeLabel_fs",
]],
PhaseProjectionND.NAME : [
{ "key": ["bert"], "family": ["all-MiniLM-L6-v2", "specter2"], "max_length": [256, 512] },
{ "key": ["lsa"], "num_dims": [50, 100, 200] }
],
PhaseProjection2D.NAME : [
{ "key": ["umap"], "n_neighbors" : [10, 20, 50], "min_dist" : [0.1, 0.25, 0.5],
"metric" : ["euclidean"] },
{ "key": ["tsne"], "perplexity" : [30, 50] }
]
}
""
param_iterator = GridIterator(params_dict=params)
""
param_iterator.params_frame.shape
""
param_iterator.params_frame
.. raw:: html
<div class="output_subarea output_html rendered_html output_result">
<div>
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
</style>
<table border="1" class="dataframe">
<thead>
<tr style="text-align: right;">
<th></th>
<th>id</th>
<th>robustseed</th>
<th>authors__filter_min_score</th>
<th>filter_min_score</th>
<th>words__column_names</th>
<th>projection_nD</th>
<th>projection_2D</th>
<th>selected</th>
</tr>
</thead>
<tbody>
<tr>
<th>0</th>
<td>5acf1c353d2c392a1729</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_keyword_s, en_domainAllCodeLabel_fs, en_ab...</td>
<td>{'key': 'bert', 'family': 'all-MiniLM-L6-v2', ...</td>
<td>{'key': 'umap', 'n_neighbors': 10, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>1</th>
<td>da34d4237bc01ee70684</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_keyword_s, en_domainAllCodeLabel_fs, en_ab...</td>
<td>{'key': 'bert', 'family': 'all-MiniLM-L6-v2', ...</td>
<td>{'key': 'umap', 'n_neighbors': 10, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>2</th>
<td>a862f98f80e4912e4e08</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_keyword_s, en_domainAllCodeLabel_fs, en_ab...</td>
<td>{'key': 'bert', 'family': 'all-MiniLM-L6-v2', ...</td>
<td>{'key': 'umap', 'n_neighbors': 10, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>3</th>
<td>a3cd00f35555d3301dab</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_keyword_s, en_domainAllCodeLabel_fs, en_ab...</td>
<td>{'key': 'bert', 'family': 'all-MiniLM-L6-v2', ...</td>
<td>{'key': 'umap', 'n_neighbors': 20, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>4</th>
<td>7e3fa98210bea7d56da4</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_keyword_s, en_domainAllCodeLabel_fs, en_ab...</td>
<td>{'key': 'bert', 'family': 'all-MiniLM-L6-v2', ...</td>
<td>{'key': 'umap', 'n_neighbors': 20, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>...</th>
<td>...</td>
<td>...</td>
<td>...</td>
<td>...</td>
<td>...</td>
<td>...</td>
<td>...</td>
<td>...</td>
</tr>
<tr>
<th>149</th>
<td>3bc0eaea7e130fc4bcef</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_abstract_s, en_title_s, en_keyword_s, en_d...</td>
<td>{'key': 'lsa', 'num_dims': 200}</td>
<td>{'key': 'umap', 'n_neighbors': 50, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>150</th>
<td>f702093065fd195285b0</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_abstract_s, en_title_s, en_keyword_s, en_d...</td>
<td>{'key': 'lsa', 'num_dims': 200}</td>
<td>{'key': 'umap', 'n_neighbors': 50, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>151</th>
<td>f3c02394b1531b5bd3ae</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_abstract_s, en_title_s, en_keyword_s, en_d...</td>
<td>{'key': 'lsa', 'num_dims': 200}</td>
<td>{'key': 'umap', 'n_neighbors': 50, 'min_dist':...</td>
<td>False</td>
</tr>
<tr>
<th>152</th>
<td>1c855fd02dab6b8f60aa</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_abstract_s, en_title_s, en_keyword_s, en_d...</td>
<td>{'key': 'lsa', 'num_dims': 200}</td>
<td>{'key': 'tsne', 'perplexity': 30}</td>
<td>False</td>
</tr>
<tr>
<th>153</th>
<td>83b2e801649d7f3bf7f8</td>
<td>0</td>
<td>4</td>
<td>6</td>
<td>[en_abstract_s, en_title_s, en_keyword_s, en_d...</td>
<td>{'key': 'lsa', 'num_dims': 200}</td>
<td>{'key': 'tsne', 'perplexity': 50}</td>
<td>False</td>
</tr>
</tbody>
</table>
<p>154 rows × 8 columns</p>
</div>
</div>
<br />
<br />
.. GENERATED FROM PYTHON SOURCE LINES 74-79
Run Experiment
--------------------------------------
We will run the expriment using `cartodata.pipeline.PipelineExperiment`.
.. GENERATED FROM PYTHON SOURCE LINES 79-84
.. code-block:: Python
from cartodata.pipeline.experiment import PipelineExperiment # noqa
help(PipelineExperiment)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Help on class PipelineExperiment in module cartodata.pipeline.experiment:
class PipelineExperiment(cartodata.model_selection.experiment.BaseExperiment)
| PipelineExperiment(dataset_name, dataset_version, top_dir, conf_dir, input_dir, nature, source, selector, score_list=None, **score_params)
|
| Method resolution order:
| PipelineExperiment
| cartodata.model_selection.experiment.BaseExperiment
| builtins.object
|
| Methods defined here:
|
| run_steps(self, next_params)
|
| ----------------------------------------------------------------------
| Methods inherited from cartodata.model_selection.experiment.BaseExperiment:
|
| __init__(self, dataset_name, dataset_version, top_dir, conf_dir, input_dir, nature, source, selector, score_list=None, **score_params)
| Parameters
| ----------
| nature: str
| source: str
| top_dir: str or Path
| the top directory inside which experiment artifacts will be
| generated
| selector: cartodata.model_selection.selector.ParameterSelector
| the selector that provides the next set of parameters for the
| experiment
| score_list: list of cartodata.model_selection.evaluators.Evaluators,
| default=None
| score_params: kwargs
| The parameters that will be used by evaluators should be specified
| as kwargs in the format ``evaluator_key__parameter_name=value``
| For example ``name_list`` parameter for ``final_score`` should be
| specified as:
|
| ``final_score__name_list=["neighbors_nd_articles_authors",
| "neighbors_2d_articles_authors",
| "clu_score"]
|
| add_2D_scores(self, all_natures, dir_mat, key_nD, key_2D, dir_nD, dir_2D, plots=None, run_params={})
|
| add_clustering_scores(self, all_natures, cluster_natures, key_nD, key_2D, dir_mat, dir_nD, dir_2D, dir_clus, words_index, plots=None, run_params={})
|
| add_nD_scores(self, key_nD=None, dir_mat=None, dir_nD=None, run_params={})
|
| add_plots_to_scoring(self, plots, natures)
|
| add_post_scores(self, dir_clus, run_params={})
|
| finish_iteration(self, next_params, dir_clus)
|
| persist_result(self, result, phase, phase_result, dump_dir)
|
| run(self, nbmax)
| Runs the experiment for the specified `nbmax` number of times.
|
| :param nbmax: maximum number of experiments to run.
|
| save_plots(self, natures, matrices, dir_dump, title_parts, annotations=None, annotation_mat=None, file_ext='.png')
|
| ----------------------------------------------------------------------
| Data descriptors inherited from cartodata.model_selection.experiment.BaseExperiment:
|
| __dict__
| dictionary for instance variables (if defined)
|
| __weakref__
| list of weak references to the object (if defined)
.. GENERATED FROM PYTHON SOURCE LINES 85-86
All possible scoring classes are in the `cartodata.model_selection.scoring` module. We will run scoring for each, so we will import them all.
.. GENERATED FROM PYTHON SOURCE LINES 86-92
.. code-block:: Python
from cartodata.model_selection.scoring import (
NeighborsND, Neighbors2D, Comparative, TrustworthinessSklearn, TrustworthinessUmap,
Clustering, FinalScore
)
.. GENERATED FROM PYTHON SOURCE LINES 93-101
We need to specify which scores to calculate to the experiment using the parameter `score_list`. If we do not specify, the experiment will evaluate scores for all available scoring classes.
We will do it explicitly and specify all classes defined in `cartodata.model_selection.scoring` module.
It is possible to specify parameters for score classes as keyword arguments.
For example, if `cartodata.model_selection.scoring.FinalScore` is specified in the `score_list`, the experiment calculates an aggregated score taking average of all scores at the end of each run. If instead of all scores, we want a subset of scores to be included in the average, we can specify it using `final_score__name_list`. For each scoring class, we should name the parameter in the format `scoring_KEY__scoring_parameter`.
.. GENERATED FROM PYTHON SOURCE LINES 101-122
.. code-block:: Python
experiment = PipelineExperiment(
"lisn", "2022.11.15.1", TOP_DIR, CONF_DIR, INPUT_DIR,
NATURE, SOURCE, param_iterator,
score_list=[NeighborsND,
Neighbors2D,
Comparative,
TrustworthinessSklearn,
# TrustworthinessUmap,
Clustering,
FinalScore],
final_score__name_list=[
PhaseProjectionND.prefix("neighbors_articles_authors"),
PhaseProjection2D.prefix("neighbors_articles_authors"),
PhaseClustering.prefix("clu_score")
],
neighbors__recompute=True,
neighbors__min_score=30,
trustworthiness_sklearn__n_neighbors=10
)
.. GENERATED FROM PYTHON SOURCE LINES 123-124
Now we will run the experiment for 3 different set of parameters.
.. GENERATED FROM PYTHON SOURCE LINES 124-127
.. code-block:: Python
results = experiment.run(3)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
Downloading data from https://zenodo.org/records/7323538/files/lisn_2000_2022.csv (6.3 MB)
file_sizes: 0%| | 0.00/6.59M [00:00<?, ?B/s]
file_sizes: 63%|████████████████▍ | 4.16M/6.59M [00:00<00:00, 39.7MB/s]
file_sizes: 100%|██████████████████████████| 6.59M/6.59M [00:00<00:00, 53.4MB/s]
Successfully downloaded file to /builds/2mk6rsew/0/hgozukan/cartolabe-data/experiment_pipeline_lisn/lisn/2022.11.15.1/lisn_2000_2022.csv
Processing batches: 0%| | 0/427 [00:00<?, ?it/s]
Processing batches: 0%| | 1/427 [00:03<21:37, 3.05s/it]
Processing batches: 0%| | 2/427 [00:05<17:03, 2.41s/it]
Processing batches: 1%| | 3/427 [00:07<18:32, 2.62s/it]
Processing batches: 1%| | 4/427 [00:11<21:59, 3.12s/it]
Processing batches: 1%| | 5/427 [00:14<21:00, 2.99s/it]
Processing batches: 1%|▏ | 6/427 [00:16<18:37, 2.65s/it]
Processing batches: 2%|▏ | 7/427 [00:18<16:55, 2.42s/it]
Processing batches: 2%|▏ | 8/427 [00:20<15:50, 2.27s/it]
Processing batches: 2%|▏ | 9/427 [00:22<16:18, 2.34s/it]
Processing batches: 2%|▏ | 10/427 [00:24<14:40, 2.11s/it]
Processing batches: 3%|▎ | 11/427 [00:25<13:01, 1.88s/it]
Processing batches: 3%|▎ | 12/427 [00:28<15:27, 2.23s/it]
Processing batches: 3%|▎ | 13/427 [00:31<15:15, 2.21s/it]
Processing batches: 3%|▎ | 14/427 [00:34<18:14, 2.65s/it]
Processing batches: 4%|▎ | 15/427 [00:36<16:45, 2.44s/it]
Processing batches: 4%|▎ | 16/427 [00:41<20:50, 3.04s/it]
Processing batches: 4%|▍ | 17/427 [00:42<17:31, 2.56s/it]
Processing batches: 4%|▍ | 18/427 [00:43<15:02, 2.21s/it]
Processing batches: 4%|▍ | 19/427 [00:46<15:58, 2.35s/it]
Processing batches: 5%|▍ | 20/427 [00:50<19:56, 2.94s/it]
Processing batches: 5%|▍ | 21/427 [00:54<20:20, 3.01s/it]
Processing batches: 5%|▌ | 22/427 [00:55<17:40, 2.62s/it]
Processing batches: 5%|▌ | 23/427 [00:58<17:31, 2.60s/it]
Processing batches: 6%|▌ | 24/427 [01:01<18:07, 2.70s/it]
Processing batches: 6%|▌ | 25/427 [01:03<16:47, 2.51s/it]
Processing batches: 6%|▌ | 26/427 [01:07<20:28, 3.06s/it]
Processing batches: 6%|▋ | 27/427 [01:09<17:47, 2.67s/it]
Processing batches: 7%|▋ | 28/427 [01:11<16:05, 2.42s/it]
Processing batches: 7%|▋ | 29/427 [01:13<15:01, 2.26s/it]
Processing batches: 7%|▋ | 30/427 [01:15<15:48, 2.39s/it]
Processing batches: 7%|▋ | 31/427 [01:17<14:41, 2.23s/it]
Processing batches: 7%|▋ | 32/427 [01:22<19:05, 2.90s/it]
Processing batches: 8%|▊ | 33/427 [01:26<21:39, 3.30s/it]
Processing batches: 8%|▊ | 34/427 [01:30<23:37, 3.61s/it]
Processing batches: 8%|▊ | 35/427 [01:32<19:50, 3.04s/it]
Processing batches: 8%|▊ | 36/427 [01:34<17:11, 2.64s/it]
Processing batches: 9%|▊ | 37/427 [01:38<20:25, 3.14s/it]
Processing batches: 9%|▉ | 38/427 [01:41<19:50, 3.06s/it]
Processing batches: 9%|▉ | 39/427 [01:42<16:20, 2.53s/it]
Processing batches: 9%|▉ | 40/427 [01:44<15:33, 2.41s/it]
Processing batches: 10%|▉ | 41/427 [01:46<14:24, 2.24s/it]
Processing batches: 10%|▉ | 42/427 [01:47<12:30, 1.95s/it]
Processing batches: 10%|█ | 43/427 [01:52<17:07, 2.68s/it]
Processing batches: 10%|█ | 44/427 [01:56<20:13, 3.17s/it]
Processing batches: 11%|█ | 45/427 [01:58<17:25, 2.74s/it]
Processing batches: 11%|█ | 46/427 [02:00<15:23, 2.42s/it]
Processing batches: 11%|█ | 47/427 [02:03<16:45, 2.65s/it]
Processing batches: 11%|█ | 48/427 [02:07<19:53, 3.15s/it]
Processing batches: 11%|█▏ | 49/427 [02:09<17:02, 2.71s/it]
Processing batches: 12%|█▏ | 50/427 [02:12<18:46, 2.99s/it]
Processing batches: 12%|█▏ | 51/427 [02:14<17:03, 2.72s/it]
Processing batches: 12%|█▏ | 52/427 [02:17<16:09, 2.59s/it]
Processing batches: 12%|█▏ | 53/427 [02:19<15:14, 2.45s/it]
Processing batches: 13%|█▎ | 54/427 [02:21<14:15, 2.29s/it]
Processing batches: 13%|█▎ | 55/427 [02:23<13:44, 2.22s/it]
Processing batches: 13%|█▎ | 56/427 [02:24<12:33, 2.03s/it]
Processing batches: 13%|█▎ | 57/427 [02:29<16:40, 2.70s/it]
Processing batches: 14%|█▎ | 58/427 [02:30<14:40, 2.39s/it]
Processing batches: 14%|█▍ | 59/427 [02:32<14:12, 2.32s/it]
Processing batches: 14%|█▍ | 60/427 [02:36<16:45, 2.74s/it]
Processing batches: 14%|█▍ | 61/427 [02:41<19:37, 3.22s/it]
Processing batches: 15%|█▍ | 62/427 [02:42<16:54, 2.78s/it]
Processing batches: 15%|█▍ | 63/427 [02:45<15:58, 2.63s/it]
Processing batches: 15%|█▍ | 64/427 [02:48<17:39, 2.92s/it]
Processing batches: 15%|█▌ | 65/427 [02:51<17:03, 2.83s/it]
Processing batches: 15%|█▌ | 66/427 [02:53<15:36, 2.59s/it]
Processing batches: 16%|█▌ | 67/427 [02:57<18:48, 3.13s/it]
Processing batches: 16%|█▌ | 68/427 [02:59<16:53, 2.82s/it]
Processing batches: 16%|█▌ | 69/427 [03:01<14:37, 2.45s/it]
Processing batches: 16%|█▋ | 70/427 [03:05<17:51, 3.00s/it]
Processing batches: 17%|█▋ | 71/427 [03:08<17:10, 2.89s/it]
Processing batches: 17%|█▋ | 72/427 [03:12<19:26, 3.29s/it]
Processing batches: 17%|█▋ | 73/427 [03:13<15:38, 2.65s/it]
Processing batches: 17%|█▋ | 74/427 [03:16<15:12, 2.58s/it]
Processing batches: 18%|█▊ | 75/427 [03:17<12:54, 2.20s/it]
Processing batches: 18%|█▊ | 76/427 [03:19<12:51, 2.20s/it]
Processing batches: 18%|█▊ | 77/427 [03:22<14:04, 2.41s/it]
Processing batches: 18%|█▊ | 78/427 [03:25<14:19, 2.46s/it]
Processing batches: 19%|█▊ | 79/427 [03:27<13:39, 2.35s/it]
Processing batches: 19%|█▊ | 80/427 [03:29<13:04, 2.26s/it]
Processing batches: 19%|█▉ | 81/427 [03:30<11:41, 2.03s/it]
Processing batches: 19%|█▉ | 82/427 [03:32<11:23, 1.98s/it]
Processing batches: 19%|█▉ | 83/427 [03:33<09:52, 1.72s/it]
Processing batches: 20%|█▉ | 84/427 [03:36<10:49, 1.89s/it]
Processing batches: 20%|█▉ | 85/427 [03:37<10:27, 1.83s/it]
Processing batches: 20%|██ | 86/427 [03:40<12:50, 2.26s/it]
Processing batches: 20%|██ | 87/427 [03:43<13:12, 2.33s/it]
Processing batches: 21%|██ | 88/427 [03:45<13:28, 2.38s/it]
Processing batches: 21%|██ | 89/427 [03:47<12:12, 2.17s/it]
Processing batches: 21%|██ | 90/427 [03:51<15:51, 2.82s/it]
Processing batches: 21%|██▏ | 91/427 [03:55<16:47, 3.00s/it]
Processing batches: 22%|██▏ | 92/427 [03:58<16:10, 2.90s/it]
Processing batches: 22%|██▏ | 93/427 [04:00<15:12, 2.73s/it]
Processing batches: 22%|██▏ | 94/427 [04:02<14:53, 2.68s/it]
Processing batches: 22%|██▏ | 95/427 [04:05<13:50, 2.50s/it]
Processing batches: 22%|██▏ | 96/427 [04:06<12:38, 2.29s/it]
Processing batches: 23%|██▎ | 97/427 [04:10<14:05, 2.56s/it]
Processing batches: 23%|██▎ | 98/427 [04:12<14:06, 2.57s/it]
Processing batches: 23%|██▎ | 99/427 [04:14<13:07, 2.40s/it]
Processing batches: 23%|██▎ | 100/427 [04:17<13:03, 2.39s/it]
Processing batches: 24%|██▎ | 101/427 [04:19<13:27, 2.48s/it]
Processing batches: 24%|██▍ | 102/427 [04:24<16:29, 3.04s/it]
Processing batches: 24%|██▍ | 103/427 [04:26<14:44, 2.73s/it]
Processing batches: 24%|██▍ | 104/427 [04:28<14:35, 2.71s/it]
Processing batches: 25%|██▍ | 105/427 [04:30<12:21, 2.30s/it]
Processing batches: 25%|██▍ | 106/427 [04:31<10:39, 1.99s/it]
Processing batches: 25%|██▌ | 107/427 [04:33<11:38, 2.18s/it]
Processing batches: 25%|██▌ | 108/427 [04:38<14:49, 2.79s/it]
Processing batches: 26%|██▌ | 109/427 [04:42<17:04, 3.22s/it]
Processing batches: 26%|██▌ | 110/427 [04:45<16:06, 3.05s/it]
Processing batches: 26%|██▌ | 111/427 [04:47<15:04, 2.86s/it]
Processing batches: 26%|██▌ | 112/427 [04:51<16:57, 3.23s/it]
Processing batches: 26%|██▋ | 113/427 [04:55<18:31, 3.54s/it]
Processing batches: 27%|██▋ | 114/427 [04:59<19:24, 3.72s/it]
Processing batches: 27%|██▋ | 115/427 [05:04<20:02, 3.86s/it]
Processing batches: 27%|██▋ | 116/427 [05:08<20:42, 4.00s/it]
Processing batches: 27%|██▋ | 117/427 [05:12<20:46, 4.02s/it]
Processing batches: 28%|██▊ | 118/427 [05:16<21:14, 4.12s/it]
Processing batches: 28%|██▊ | 119/427 [05:19<19:16, 3.75s/it]
Processing batches: 28%|██▊ | 120/427 [05:21<16:27, 3.22s/it]
Processing batches: 28%|██▊ | 121/427 [05:23<13:41, 2.68s/it]
Processing batches: 29%|██▊ | 122/427 [05:25<12:20, 2.43s/it]
Processing batches: 29%|██▉ | 123/427 [05:28<14:21, 2.83s/it]
Processing batches: 29%|██▉ | 124/427 [05:29<11:45, 2.33s/it]
Processing batches: 29%|██▉ | 125/427 [05:31<11:00, 2.19s/it]
Processing batches: 30%|██▉ | 126/427 [05:36<14:09, 2.82s/it]
Processing batches: 30%|██▉ | 127/427 [05:38<14:10, 2.84s/it]
Processing batches: 30%|██▉ | 128/427 [05:43<16:22, 3.29s/it]
Processing batches: 30%|███ | 129/427 [05:45<14:11, 2.86s/it]
Processing batches: 30%|███ | 130/427 [05:47<13:22, 2.70s/it]
Processing batches: 31%|███ | 131/427 [05:51<15:41, 3.18s/it]
Processing batches: 31%|███ | 132/427 [05:54<14:38, 2.98s/it]
Processing batches: 31%|███ | 133/427 [05:56<13:49, 2.82s/it]
Processing batches: 31%|███▏ | 134/427 [05:58<11:27, 2.35s/it]
Processing batches: 32%|███▏ | 135/427 [06:02<14:12, 2.92s/it]
Processing batches: 32%|███▏ | 136/427 [06:03<12:02, 2.48s/it]
Processing batches: 32%|███▏ | 137/427 [06:07<14:10, 2.93s/it]
Processing batches: 32%|███▏ | 138/427 [06:10<13:58, 2.90s/it]
Processing batches: 33%|███▎ | 139/427 [06:14<15:46, 3.28s/it]
Processing batches: 33%|███▎ | 140/427 [06:18<17:01, 3.56s/it]
Processing batches: 33%|███▎ | 141/427 [06:20<14:42, 3.09s/it]
Processing batches: 33%|███▎ | 142/427 [06:23<13:19, 2.80s/it]
Processing batches: 33%|███▎ | 143/427 [06:27<15:01, 3.17s/it]
Processing batches: 34%|███▎ | 144/427 [06:31<16:05, 3.41s/it]
Processing batches: 34%|███▍ | 145/427 [06:35<17:10, 3.66s/it]
Processing batches: 34%|███▍ | 146/427 [06:39<17:55, 3.83s/it]
Processing batches: 34%|███▍ | 147/427 [06:41<14:45, 3.16s/it]
Processing batches: 35%|███▍ | 148/427 [06:45<16:21, 3.52s/it]
Processing batches: 35%|███▍ | 149/427 [06:47<14:07, 3.05s/it]
Processing batches: 35%|███▌ | 150/427 [06:49<12:38, 2.74s/it]
Processing batches: 35%|███▌ | 151/427 [06:50<10:53, 2.37s/it]
Processing batches: 36%|███▌ | 152/427 [06:53<10:25, 2.27s/it]
Processing batches: 36%|███▌ | 153/427 [06:55<11:10, 2.45s/it]
Processing batches: 36%|███▌ | 154/427 [07:00<13:57, 3.07s/it]
Processing batches: 36%|███▋ | 155/427 [07:02<12:57, 2.86s/it]
Processing batches: 37%|███▋ | 156/427 [07:07<14:50, 3.29s/it]
Processing batches: 37%|███▋ | 157/427 [07:11<16:07, 3.58s/it]
Processing batches: 37%|███▋ | 158/427 [07:15<16:54, 3.77s/it]
Processing batches: 37%|███▋ | 159/427 [07:19<17:21, 3.89s/it]
Processing batches: 37%|███▋ | 160/427 [07:23<17:00, 3.82s/it]
Processing batches: 38%|███▊ | 161/427 [07:26<15:50, 3.57s/it]
Processing batches: 38%|███▊ | 162/427 [07:29<15:41, 3.55s/it]
Processing batches: 38%|███▊ | 163/427 [07:31<13:24, 3.05s/it]
Processing batches: 38%|███▊ | 164/427 [07:35<14:44, 3.36s/it]
Processing batches: 39%|███▊ | 165/427 [07:39<14:44, 3.38s/it]
Processing batches: 39%|███▉ | 166/427 [07:43<15:36, 3.59s/it]
Processing batches: 39%|███▉ | 167/427 [07:47<15:43, 3.63s/it]
Processing batches: 39%|███▉ | 168/427 [07:51<16:39, 3.86s/it]
Processing batches: 40%|███▉ | 169/427 [07:55<16:18, 3.79s/it]
Processing batches: 40%|███▉ | 170/427 [07:59<16:44, 3.91s/it]
Processing batches: 40%|████ | 171/427 [08:01<14:54, 3.50s/it]
Processing batches: 40%|████ | 172/427 [08:04<13:41, 3.22s/it]
Processing batches: 41%|████ | 173/427 [08:08<14:40, 3.47s/it]
Processing batches: 41%|████ | 174/427 [08:10<12:50, 3.04s/it]
Processing batches: 41%|████ | 175/427 [08:13<13:21, 3.18s/it]
Processing batches: 41%|████ | 176/427 [08:18<14:53, 3.56s/it]
Processing batches: 41%|████▏ | 177/427 [08:22<15:00, 3.60s/it]
Processing batches: 42%|████▏ | 178/427 [08:25<14:22, 3.46s/it]
Processing batches: 42%|████▏ | 179/427 [08:26<11:31, 2.79s/it]
Processing batches: 42%|████▏ | 180/427 [08:28<11:10, 2.72s/it]
Processing batches: 42%|████▏ | 181/427 [08:33<13:08, 3.20s/it]
Processing batches: 43%|████▎ | 182/427 [08:35<12:07, 2.97s/it]
Processing batches: 43%|████▎ | 183/427 [08:40<13:39, 3.36s/it]
Processing batches: 43%|████▎ | 184/427 [08:41<11:55, 2.94s/it]
Processing batches: 43%|████▎ | 185/427 [08:45<12:19, 3.06s/it]
Processing batches: 44%|████▎ | 186/427 [08:48<12:30, 3.11s/it]
Processing batches: 44%|████▍ | 187/427 [08:50<11:13, 2.81s/it]
Processing batches: 44%|████▍ | 188/427 [08:53<11:06, 2.79s/it]
Processing batches: 44%|████▍ | 189/427 [08:57<12:59, 3.28s/it]
Processing batches: 44%|████▍ | 190/427 [09:02<14:20, 3.63s/it]
Processing batches: 45%|████▍ | 191/427 [09:06<14:55, 3.79s/it]
Processing batches: 45%|████▍ | 192/427 [09:10<15:26, 3.94s/it]
Processing batches: 45%|████▌ | 193/427 [09:14<15:44, 4.04s/it]
Processing batches: 45%|████▌ | 194/427 [09:19<15:54, 4.09s/it]
Processing batches: 46%|████▌ | 195/427 [09:21<13:39, 3.53s/it]
Processing batches: 46%|████▌ | 196/427 [09:23<12:19, 3.20s/it]
Processing batches: 46%|████▌ | 197/427 [09:28<13:35, 3.55s/it]
Processing batches: 46%|████▋ | 198/427 [09:32<14:24, 3.78s/it]
Processing batches: 47%|████▋ | 199/427 [09:34<11:54, 3.13s/it]
Processing batches: 47%|████▋ | 200/427 [09:35<10:18, 2.72s/it]
Processing batches: 47%|████▋ | 201/427 [09:40<11:52, 3.15s/it]
Processing batches: 47%|████▋ | 202/427 [09:44<13:14, 3.53s/it]
Processing batches: 48%|████▊ | 203/427 [09:45<10:53, 2.92s/it]
Processing batches: 48%|████▊ | 204/427 [09:47<09:07, 2.46s/it]
Processing batches: 48%|████▊ | 205/427 [09:50<09:49, 2.66s/it]
Processing batches: 48%|████▊ | 206/427 [09:54<11:19, 3.07s/it]
Processing batches: 48%|████▊ | 207/427 [09:56<09:32, 2.60s/it]
Processing batches: 49%|████▊ | 208/427 [09:58<09:12, 2.53s/it]
Processing batches: 49%|████▉ | 209/427 [10:01<09:51, 2.71s/it]
Processing batches: 49%|████▉ | 210/427 [10:03<09:31, 2.63s/it]
Processing batches: 49%|████▉ | 211/427 [10:08<11:09, 3.10s/it]
Processing batches: 50%|████▉ | 212/427 [10:09<09:44, 2.72s/it]
Processing batches: 50%|████▉ | 213/427 [10:11<08:24, 2.36s/it]
Processing batches: 50%|█████ | 214/427 [10:13<08:06, 2.28s/it]
Processing batches: 50%|█████ | 215/427 [10:17<10:15, 2.90s/it]
Processing batches: 51%|█████ | 216/427 [10:19<08:16, 2.35s/it]
Processing batches: 51%|█████ | 217/427 [10:20<07:04, 2.02s/it]
Processing batches: 51%|█████ | 218/427 [10:24<09:10, 2.63s/it]
Processing batches: 51%|█████▏ | 219/427 [10:27<09:39, 2.79s/it]
Processing batches: 52%|█████▏ | 220/427 [10:30<10:18, 2.99s/it]
Processing batches: 52%|█████▏ | 221/427 [10:33<10:07, 2.95s/it]
Processing batches: 52%|█████▏ | 222/427 [10:36<09:52, 2.89s/it]
Processing batches: 52%|█████▏ | 223/427 [10:40<11:09, 3.28s/it]
Processing batches: 52%|█████▏ | 224/427 [10:45<12:16, 3.63s/it]
Processing batches: 53%|█████▎ | 225/427 [10:49<12:56, 3.84s/it]
Processing batches: 53%|█████▎ | 226/427 [10:53<13:25, 4.01s/it]
Processing batches: 53%|█████▎ | 227/427 [10:57<12:44, 3.82s/it]
Processing batches: 53%|█████▎ | 228/427 [11:01<13:12, 3.98s/it]
Processing batches: 54%|█████▎ | 229/427 [11:05<13:27, 4.08s/it]
Processing batches: 54%|█████▍ | 230/427 [11:10<13:33, 4.13s/it]
Processing batches: 54%|█████▍ | 231/427 [11:14<13:36, 4.17s/it]
Processing batches: 54%|█████▍ | 232/427 [11:16<11:17, 3.48s/it]
Processing batches: 55%|█████▍ | 233/427 [11:18<09:44, 3.01s/it]
Processing batches: 55%|█████▍ | 234/427 [11:22<11:10, 3.47s/it]
Processing batches: 55%|█████▌ | 235/427 [11:26<11:45, 3.68s/it]
Processing batches: 55%|█████▌ | 236/427 [11:30<11:21, 3.57s/it]
Processing batches: 56%|█████▌ | 237/427 [11:34<11:41, 3.69s/it]
Processing batches: 56%|█████▌ | 238/427 [11:37<11:12, 3.56s/it]
Processing batches: 56%|█████▌ | 239/427 [11:39<10:07, 3.23s/it]
Processing batches: 56%|█████▌ | 240/427 [11:43<10:12, 3.27s/it]
Processing batches: 56%|█████▋ | 241/427 [11:46<09:33, 3.08s/it]
Processing batches: 57%|█████▋ | 242/427 [11:50<10:44, 3.48s/it]
Processing batches: 57%|█████▋ | 243/427 [11:52<09:10, 2.99s/it]
Processing batches: 57%|█████▋ | 244/427 [11:56<10:23, 3.41s/it]
Processing batches: 57%|█████▋ | 245/427 [11:58<08:43, 2.88s/it]
Processing batches: 58%|█████▊ | 246/427 [12:00<08:15, 2.74s/it]
Processing batches: 58%|█████▊ | 247/427 [12:01<06:49, 2.27s/it]
Processing batches: 58%|█████▊ | 248/427 [12:03<06:04, 2.04s/it]
Processing batches: 58%|█████▊ | 249/427 [12:06<06:37, 2.23s/it]
Processing batches: 59%|█████▊ | 250/427 [12:08<06:35, 2.24s/it]
Processing batches: 59%|█████▉ | 251/427 [12:12<08:15, 2.81s/it]
Processing batches: 59%|█████▉ | 252/427 [12:16<09:12, 3.15s/it]
Processing batches: 59%|█████▉ | 253/427 [12:20<09:56, 3.43s/it]
Processing batches: 59%|█████▉ | 254/427 [12:22<08:28, 2.94s/it]
Processing batches: 60%|█████▉ | 255/427 [12:25<08:28, 2.96s/it]
Processing batches: 60%|█████▉ | 256/427 [12:27<07:36, 2.67s/it]
Processing batches: 60%|██████ | 257/427 [12:29<07:20, 2.59s/it]
Processing batches: 60%|██████ | 258/427 [12:32<07:27, 2.65s/it]
Processing batches: 61%|██████ | 259/427 [12:35<07:24, 2.64s/it]
Processing batches: 61%|██████ | 260/427 [12:39<08:33, 3.07s/it]
Processing batches: 61%|██████ | 261/427 [12:43<09:35, 3.47s/it]
Processing batches: 61%|██████▏ | 262/427 [12:47<10:04, 3.67s/it]
Processing batches: 62%|██████▏ | 263/427 [12:51<10:30, 3.84s/it]
Processing batches: 62%|██████▏ | 264/427 [12:53<08:13, 3.03s/it]
Processing batches: 62%|██████▏ | 265/427 [12:55<07:38, 2.83s/it]
Processing batches: 62%|██████▏ | 266/427 [12:57<06:45, 2.52s/it]
Processing batches: 63%|██████▎ | 267/427 [13:01<08:16, 3.11s/it]
Processing batches: 63%|██████▎ | 268/427 [13:03<07:16, 2.75s/it]
Processing batches: 63%|██████▎ | 269/427 [13:07<08:21, 3.17s/it]
Processing batches: 63%|██████▎ | 270/427 [13:10<08:07, 3.11s/it]
Processing batches: 63%|██████▎ | 271/427 [13:14<08:55, 3.43s/it]
Processing batches: 64%|██████▎ | 272/427 [13:19<09:28, 3.67s/it]
Processing batches: 64%|██████▍ | 273/427 [13:21<08:45, 3.41s/it]
Processing batches: 64%|██████▍ | 274/427 [13:26<09:15, 3.63s/it]
Processing batches: 64%|██████▍ | 275/427 [13:30<09:38, 3.80s/it]
Processing batches: 65%|██████▍ | 276/427 [13:32<08:15, 3.28s/it]
Processing batches: 65%|██████▍ | 277/427 [13:34<07:06, 2.84s/it]
Processing batches: 65%|██████▌ | 278/427 [13:35<06:16, 2.53s/it]
Processing batches: 65%|██████▌ | 279/427 [13:37<05:21, 2.17s/it]
Processing batches: 66%|██████▌ | 280/427 [13:41<06:48, 2.78s/it]
Processing batches: 66%|██████▌ | 281/427 [13:42<05:38, 2.32s/it]
Processing batches: 66%|██████▌ | 282/427 [13:44<04:57, 2.05s/it]
Processing batches: 66%|██████▋ | 283/427 [13:46<05:03, 2.10s/it]
Processing batches: 67%|██████▋ | 284/427 [13:47<04:27, 1.87s/it]
Processing batches: 67%|██████▋ | 285/427 [13:52<06:10, 2.61s/it]
Processing batches: 67%|██████▋ | 286/427 [13:54<05:52, 2.50s/it]
Processing batches: 67%|██████▋ | 287/427 [13:55<05:06, 2.19s/it]
Processing batches: 67%|██████▋ | 288/427 [14:00<06:30, 2.81s/it]
Processing batches: 68%|██████▊ | 289/427 [14:04<07:21, 3.20s/it]
Processing batches: 68%|██████▊ | 290/427 [14:06<06:55, 3.04s/it]
Processing batches: 68%|██████▊ | 291/427 [14:11<07:46, 3.43s/it]
Processing batches: 68%|██████▊ | 292/427 [14:15<08:18, 3.69s/it]
Processing batches: 69%|██████▊ | 293/427 [14:20<08:51, 3.97s/it]
Processing batches: 69%|██████▉ | 294/427 [14:21<07:24, 3.34s/it]
Processing batches: 69%|██████▉ | 295/427 [14:24<06:53, 3.13s/it]
Processing batches: 69%|██████▉ | 296/427 [14:28<07:19, 3.36s/it]
Processing batches: 70%|██████▉ | 297/427 [14:31<07:15, 3.35s/it]
Processing batches: 70%|██████▉ | 298/427 [14:33<06:21, 2.96s/it]
Processing batches: 70%|███████ | 299/427 [14:35<05:38, 2.64s/it]
Processing batches: 70%|███████ | 300/427 [14:37<05:17, 2.50s/it]
Processing batches: 70%|███████ | 301/427 [14:41<06:03, 2.89s/it]
Processing batches: 71%|███████ | 302/427 [14:43<05:31, 2.65s/it]
Processing batches: 71%|███████ | 303/427 [14:45<05:10, 2.50s/it]
Processing batches: 71%|███████ | 304/427 [14:48<05:10, 2.53s/it]
Processing batches: 71%|███████▏ | 305/427 [14:51<05:23, 2.65s/it]
Processing batches: 72%|███████▏ | 306/427 [14:54<05:17, 2.63s/it]
Processing batches: 72%|███████▏ | 307/427 [14:58<06:19, 3.16s/it]
Processing batches: 72%|███████▏ | 308/427 [15:01<06:09, 3.11s/it]
Processing batches: 72%|███████▏ | 309/427 [15:05<06:44, 3.43s/it]
Processing batches: 73%|███████▎ | 310/427 [15:08<06:35, 3.38s/it]
Processing batches: 73%|███████▎ | 311/427 [15:10<05:33, 2.88s/it]
Processing batches: 73%|███████▎ | 312/427 [15:14<06:17, 3.28s/it]
Processing batches: 73%|███████▎ | 313/427 [15:18<06:16, 3.30s/it]
Processing batches: 74%|███████▎ | 314/427 [15:20<05:23, 2.86s/it]
Processing batches: 74%|███████▍ | 315/427 [15:24<06:07, 3.28s/it]
Processing batches: 74%|███████▍ | 316/427 [15:28<06:32, 3.54s/it]
Processing batches: 74%|███████▍ | 317/427 [15:32<06:51, 3.74s/it]
Processing batches: 74%|███████▍ | 318/427 [15:36<06:36, 3.64s/it]
Processing batches: 75%|███████▍ | 319/427 [15:38<05:51, 3.26s/it]
Processing batches: 75%|███████▍ | 320/427 [15:40<05:15, 2.95s/it]
Processing batches: 75%|███████▌ | 321/427 [15:44<05:53, 3.34s/it]
Processing batches: 75%|███████▌ | 322/427 [15:46<04:58, 2.84s/it]
Processing batches: 76%|███████▌ | 323/427 [15:49<04:47, 2.77s/it]
Processing batches: 76%|███████▌ | 324/427 [15:53<05:26, 3.17s/it]
Processing batches: 76%|███████▌ | 325/427 [15:56<05:10, 3.04s/it]
Processing batches: 76%|███████▋ | 326/427 [16:00<05:46, 3.43s/it]
Processing batches: 77%|███████▋ | 327/427 [16:03<05:24, 3.25s/it]
Processing batches: 77%|███████▋ | 328/427 [16:05<04:47, 2.91s/it]
Processing batches: 77%|███████▋ | 329/427 [16:06<04:00, 2.45s/it]
Processing batches: 77%|███████▋ | 330/427 [16:09<04:02, 2.50s/it]
Processing batches: 78%|███████▊ | 331/427 [16:12<04:08, 2.58s/it]
Processing batches: 78%|███████▊ | 332/427 [16:15<04:33, 2.88s/it]
Processing batches: 78%|███████▊ | 333/427 [16:19<05:08, 3.28s/it]
Processing batches: 78%|███████▊ | 334/427 [16:24<05:33, 3.58s/it]
Processing batches: 78%|███████▊ | 335/427 [16:28<05:43, 3.73s/it]
Processing batches: 79%|███████▊ | 336/427 [16:32<05:52, 3.87s/it]
Processing batches: 79%|███████▉ | 337/427 [16:35<05:38, 3.76s/it]
Processing batches: 79%|███████▉ | 338/427 [16:39<05:22, 3.62s/it]
Processing batches: 79%|███████▉ | 339/427 [16:42<05:13, 3.56s/it]
Processing batches: 80%|███████▉ | 340/427 [16:46<05:25, 3.74s/it]
Processing batches: 80%|███████▉ | 341/427 [16:50<05:28, 3.82s/it]
Processing batches: 80%|████████ | 342/427 [16:54<05:18, 3.75s/it]
Processing batches: 80%|████████ | 343/427 [16:56<04:36, 3.29s/it]
Processing batches: 81%|████████ | 344/427 [16:58<04:02, 2.92s/it]
Processing batches: 81%|████████ | 345/427 [17:00<03:33, 2.60s/it]
Processing batches: 81%|████████ | 346/427 [17:04<04:07, 3.06s/it]
Processing batches: 81%|████████▏ | 347/427 [17:07<03:54, 2.94s/it]
Processing batches: 81%|████████▏ | 348/427 [17:09<03:23, 2.58s/it]
Processing batches: 82%|████████▏ | 349/427 [17:12<03:50, 2.95s/it]
Processing batches: 82%|████████▏ | 350/427 [17:15<03:29, 2.71s/it]
Processing batches: 82%|████████▏ | 351/427 [17:18<03:54, 3.08s/it]
Processing batches: 82%|████████▏ | 352/427 [17:22<03:54, 3.13s/it]
Processing batches: 83%|████████▎ | 353/427 [17:26<04:17, 3.48s/it]
Processing batches: 83%|████████▎ | 354/427 [17:30<04:30, 3.70s/it]
Processing batches: 83%|████████▎ | 355/427 [17:34<04:37, 3.86s/it]
Processing batches: 83%|████████▎ | 356/427 [17:39<04:41, 3.96s/it]
Processing batches: 84%|████████▎ | 357/427 [17:42<04:33, 3.90s/it]
Processing batches: 84%|████████▍ | 358/427 [17:47<04:37, 4.02s/it]
Processing batches: 84%|████████▍ | 359/427 [17:48<03:47, 3.35s/it]
Processing batches: 84%|████████▍ | 360/427 [17:51<03:30, 3.14s/it]
Processing batches: 85%|████████▍ | 361/427 [17:53<03:01, 2.74s/it]
Processing batches: 85%|████████▍ | 362/427 [17:56<03:08, 2.91s/it]
Processing batches: 85%|████████▌ | 363/427 [17:58<02:50, 2.66s/it]
Processing batches: 85%|████████▌ | 364/427 [18:01<02:47, 2.66s/it]
Processing batches: 85%|████████▌ | 365/427 [18:05<03:10, 3.08s/it]
Processing batches: 86%|████████▌ | 366/427 [18:07<02:54, 2.86s/it]
Processing batches: 86%|████████▌ | 367/427 [18:10<02:51, 2.86s/it]
Processing batches: 86%|████████▌ | 368/427 [18:14<03:09, 3.21s/it]
Processing batches: 86%|████████▋ | 369/427 [18:18<03:23, 3.50s/it]
Processing batches: 87%|████████▋ | 370/427 [18:23<03:33, 3.75s/it]
Processing batches: 87%|████████▋ | 371/427 [18:27<03:37, 3.89s/it]
Processing batches: 87%|████████▋ | 372/427 [18:29<02:57, 3.23s/it]
Processing batches: 87%|████████▋ | 373/427 [18:33<03:08, 3.49s/it]
Processing batches: 88%|████████▊ | 374/427 [18:37<03:18, 3.74s/it]
Processing batches: 88%|████████▊ | 375/427 [18:40<02:54, 3.35s/it]
Processing batches: 88%|████████▊ | 376/427 [18:43<02:46, 3.27s/it]
Processing batches: 88%|████████▊ | 377/427 [18:46<02:44, 3.28s/it]
Processing batches: 89%|████████▊ | 378/427 [18:48<02:17, 2.81s/it]
Processing batches: 89%|████████▉ | 379/427 [18:50<02:05, 2.62s/it]
Processing batches: 89%|████████▉ | 380/427 [18:52<01:59, 2.55s/it]
Processing batches: 89%|████████▉ | 381/427 [18:56<02:19, 3.03s/it]
Processing batches: 89%|████████▉ | 382/427 [18:58<01:56, 2.58s/it]
Processing batches: 90%|████████▉ | 383/427 [19:02<02:15, 3.08s/it]
Processing batches: 90%|████████▉ | 384/427 [19:05<02:10, 3.03s/it]
Processing batches: 90%|█████████ | 385/427 [19:09<02:20, 3.35s/it]
Processing batches: 90%|█████████ | 386/427 [19:12<02:07, 3.12s/it]
Processing batches: 91%|█████████ | 387/427 [19:14<01:53, 2.84s/it]
Processing batches: 91%|█████████ | 388/427 [19:16<01:40, 2.58s/it]
Processing batches: 91%|█████████ | 389/427 [19:20<01:55, 3.03s/it]
Processing batches: 91%|█████████▏| 390/427 [19:22<01:36, 2.62s/it]
Processing batches: 92%|█████████▏| 391/427 [19:26<01:48, 3.01s/it]
Processing batches: 92%|█████████▏| 392/427 [19:30<01:55, 3.31s/it]
Processing batches: 92%|█████████▏| 393/427 [19:31<01:31, 2.68s/it]
Processing batches: 92%|█████████▏| 394/427 [19:35<01:44, 3.16s/it]
Processing batches: 93%|█████████▎| 395/427 [19:38<01:36, 3.03s/it]
Processing batches: 93%|█████████▎| 396/427 [19:42<01:45, 3.39s/it]
Processing batches: 93%|█████████▎| 397/427 [19:46<01:47, 3.59s/it]
Processing batches: 93%|█████████▎| 398/427 [19:48<01:33, 3.21s/it]
Processing batches: 93%|█████████▎| 399/427 [19:51<01:22, 2.95s/it]
Processing batches: 94%|█████████▎| 400/427 [19:55<01:28, 3.29s/it]
Processing batches: 94%|█████████▍| 401/427 [19:59<01:32, 3.56s/it]
Processing batches: 94%|█████████▍| 402/427 [20:03<01:32, 3.69s/it]
Processing batches: 94%|█████████▍| 403/427 [20:07<01:32, 3.84s/it]
Processing batches: 95%|█████████▍| 404/427 [20:10<01:21, 3.56s/it]
Processing batches: 95%|█████████▍| 405/427 [20:13<01:13, 3.35s/it]
Processing batches: 95%|█████████▌| 406/427 [20:15<01:04, 3.05s/it]
Processing batches: 95%|█████████▌| 407/427 [20:18<00:56, 2.84s/it]
Processing batches: 96%|█████████▌| 408/427 [20:22<01:00, 3.16s/it]
Processing batches: 96%|█████████▌| 409/427 [20:26<01:01, 3.43s/it]
Processing batches: 96%|█████████▌| 410/427 [20:30<01:01, 3.60s/it]
Processing batches: 96%|█████████▋| 411/427 [20:34<01:00, 3.78s/it]
Processing batches: 96%|█████████▋| 412/427 [20:37<00:51, 3.45s/it]
Processing batches: 97%|█████████▋| 413/427 [20:39<00:44, 3.15s/it]
Processing batches: 97%|█████████▋| 414/427 [20:41<00:38, 2.95s/it]
Processing batches: 97%|█████████▋| 415/427 [20:46<00:39, 3.31s/it]
Processing batches: 97%|█████████▋| 416/427 [20:50<00:39, 3.57s/it]
Processing batches: 98%|█████████▊| 417/427 [20:52<00:32, 3.29s/it]
Processing batches: 98%|█████████▊| 418/427 [20:55<00:27, 3.01s/it]
Processing batches: 98%|█████████▊| 419/427 [20:57<00:21, 2.68s/it]
Processing batches: 98%|█████████▊| 420/427 [20:59<00:18, 2.61s/it]
Processing batches: 99%|█████████▊| 421/427 [21:03<00:18, 3.09s/it]
Processing batches: 99%|█████████▉| 422/427 [21:07<00:16, 3.35s/it]
Processing batches: 99%|█████████▉| 423/427 [21:10<00:12, 3.16s/it]
Processing batches: 99%|█████████▉| 424/427 [21:12<00:08, 2.87s/it]
Processing batches: 100%|█████████▉| 425/427 [21:16<00:06, 3.22s/it]
Processing batches: 100%|█████████▉| 426/427 [21:20<00:03, 3.45s/it]
Processing batches: 100%|██████████| 427/427 [21:20<00:00, 2.47s/it]
Processing batches: 100%|██████████| 427/427 [21:20<00:00, 3.00s/it]
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 10, '3_2D__rmin_dist': 0.1, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.5361
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.0148
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0111
3_2D__trustworthiness_sklearn : 0.9486
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Céline Gicquel : 0.79
1 Loïc Paulevé : 0.76
2 Isabelle Guyon : 0.75
3 Nikolaus Hansen : 0.74
4 Sébastien Tixeuil : 0.70
5 Paola Tubaro : 0.69
6 Dimo Brockhoff : 0.67
7 Franck Cappello : 0.66
8 Sarah Cohen-Boulakia : 0.66
9 Cyril Furtlehner : 0.66
10 Tobias Isenberg : 0.64
11 Fatiha Zaidi : 0.64
12 Yann Ponty : 0.64
13 Albert Cohen : 0.63
14 Nathalie Pernelle : 0.62
15 Fatiha Saïs : 0.61
16 Marc Baboulin : 0.59
17 Nicolas Bredeche : 0.58
18 Guillaume Charpiat : 0.58
19 Ioana Manolescu : 0.58
20 Olivier Teytaud : 0.58
21 Marc Schoenauer : 0.57
22 Raymond Ros : 0.57
23 Chantal Reynaud : 0.57
24 Philippe Caillou : 0.57
25 Petra Isenberg : 0.56
26 Anne Auger : 0.54
27 Jean-Daniel Fekete : 0.52
28 Lonni Besançon : 0.52
29 Sylvie Boldo : 0.49
30 Claude Marché : 0.49
31 François Goasdoué : 0.48
32 Michèle Sebag : 0.48
33 Nathann Cohen : 0.48
34 Guillaume Melquiond : 0.46
35 Steven Martin : 0.45
36 Evelyne Lutton : 0.44
37 Balázs Kégl : 0.41
38 Michel Beaudouin-Lafon : 0.40
39 Olivier Chapuis : 0.39
40 Caroline Appert : 0.37
41 Alain Denise : 0.37
42 Anastasia Bezerianos : 0.34
43 Pierre Dragicevic : 0.33
44 Emmanuel Pietriga : 0.32
45 Wendy Mackay : 0.29
46 Johanne Cohen : 0.28
47 Wendy E. Mackay : 0.27
dtype: object
VALUE : 0.5361
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 François Goasdoué : -0.14
1 Anne Auger : -0.14
2 Guillaume Melquiond : -0.11
3 Evelyne Lutton : -0.10
4 Jean-Daniel Fekete : -0.08
5 Ioana Manolescu : -0.08
6 Dimo Brockhoff : -0.08
7 Céline Gicquel : -0.07
8 Sylvie Boldo : -0.06
9 Pierre Dragicevic : -0.06
10 Marc Baboulin : -0.06
11 Philippe Caillou : -0.05
12 Wendy Mackay : -0.05
13 Nicolas Bredeche : -0.05
14 Johanne Cohen : -0.04
15 Nathann Cohen : -0.04
16 Emmanuel Pietriga : -0.04
17 Claude Marché : -0.04
18 Petra Isenberg : -0.04
19 Tobias Isenberg : -0.03
20 Yann Ponty : -0.03
21 Loïc Paulevé : -0.03
22 Nikolaus Hansen : -0.03
23 Michel Beaudouin-Lafon : -0.01
24 Fatiha Saïs : -0.01
25 Olivier Chapuis : -0.01
26 Lonni Besançon : -0.01
27 Anastasia Bezerianos : 0.00
28 Olivier Teytaud : 0.00
29 Balázs Kégl : 0.01
30 Nathalie Pernelle : 0.01
31 Alain Denise : 0.01
32 Franck Cappello : 0.01
33 Cyril Furtlehner : 0.01
34 Marc Schoenauer : 0.02
35 Paola Tubaro : 0.02
36 Guillaume Charpiat : 0.02
37 Caroline Appert : 0.03
38 Steven Martin : 0.03
39 Raymond Ros : 0.04
40 Wendy E. Mackay : 0.04
41 Michèle Sebag : 0.05
42 Chantal Reynaud : 0.06
43 Albert Cohen : 0.06
44 Sébastien Tixeuil : 0.07
45 Fatiha Zaidi : 0.09
46 Isabelle Guyon : 0.09
47 Sarah Cohen-Boulakia : 0.09
dtype: object
VALUE : -0.0148
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.695745
Michèle Sebag 0.483212
Johanne Cohen 0.281395
Albert Cohen 0.625862
Wendy E. Mackay 0.269565
Philippe Caillou 0.565116
Alain Denise 0.366667
Jean-Daniel Fekete 0.520755
Emmanuel Pietriga 0.323810
Yann Ponty 0.636364
Marc Schoenauer 0.574820
Franck Cappello 0.663415
Caroline Appert 0.371739
Michel Beaudouin-Lafon 0.401235
Wendy Mackay 0.291489
Anne Auger 0.541772
Evelyne Lutton 0.443243
Pierre Dragicevic 0.334568
Ioana Manolescu 0.579268
Nikolaus Hansen 0.735802
Nicolas Bredeche 0.582353
Olivier Teytaud 0.578302
François Goasdoué 0.484906
Nathalie Pernelle 0.623529
Fatiha Saïs 0.612195
Sarah Cohen-Boulakia 0.657576
Claude Marché 0.489362
Chantal Reynaud 0.570000
Olivier Chapuis 0.390385
Steven Martin 0.451282
Fatiha Zaidi 0.637500
Balázs Kégl 0.407895
Paola Tubaro 0.689744
Raymond Ros 0.573529
Cyril Furtlehner 0.656410
Anastasia Bezerianos 0.337313
Sylvie Boldo 0.494286
Guillaume Melquiond 0.457576
Marc Baboulin 0.593333
Dimo Brockhoff 0.671795
Nathann Cohen 0.478049
Petra Isenberg 0.555140
Tobias Isenberg 0.637607
Loïc Paulevé 0.759524
Céline Gicquel 0.789474
Isabelle Guyon 0.747191
Guillaume Charpiat 0.580645
Lonni Besançon 0.518182
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.695745 0.072340
Michèle Sebag 0.483212 0.050365
Johanne Cohen 0.281395 -0.044186
Albert Cohen 0.625862 0.060345
Wendy E. Mackay 0.269565 0.039130
Philippe Caillou 0.565116 -0.053488
Alain Denise 0.366667 0.008333
Jean-Daniel Fekete 0.520755 -0.080503
Emmanuel Pietriga 0.323810 -0.039683
Yann Ponty 0.636364 -0.027273
Marc Schoenauer 0.574820 0.015827
Franck Cappello 0.663415 0.012195
Caroline Appert 0.371739 0.030435
Michel Beaudouin-Lafon 0.401235 -0.012346
Wendy Mackay 0.291489 -0.053191
Anne Auger 0.541772 -0.135443
Evelyne Lutton 0.443243 -0.097297
Pierre Dragicevic 0.334568 -0.058025
Ioana Manolescu 0.579268 -0.080488
Nikolaus Hansen 0.735802 -0.025926
Nicolas Bredeche 0.582353 -0.052941
Olivier Teytaud 0.578302 0.002830
François Goasdoué 0.484906 -0.135849
Nathalie Pernelle 0.623529 0.005882
Fatiha Saïs 0.612195 -0.009756
Sarah Cohen-Boulakia 0.657576 0.093939
Claude Marché 0.489362 -0.038298
Chantal Reynaud 0.570000 0.060000
Olivier Chapuis 0.390385 -0.007692
Steven Martin 0.451282 0.033333
Fatiha Zaidi 0.637500 0.087500
Balázs Kégl 0.407895 0.005263
Paola Tubaro 0.689744 0.017949
Raymond Ros 0.573529 0.038235
Cyril Furtlehner 0.656410 0.012821
Anastasia Bezerianos 0.337313 0.001493
Sylvie Boldo 0.494286 -0.062857
Guillaume Melquiond 0.457576 -0.106061
Marc Baboulin 0.593333 -0.057778
Dimo Brockhoff 0.671795 -0.079487
Nathann Cohen 0.478049 -0.041463
Petra Isenberg 0.555140 -0.037383
Tobias Isenberg 0.637607 -0.034188
Loïc Paulevé 0.759524 -0.026190
Céline Gicquel 0.789474 -0.068421
Isabelle Guyon 0.747191 0.091011
Guillaume Charpiat 0.580645 0.022581
Lonni Besançon 0.518182 -0.006061
Nothing in cache, initial Fitting with min_cluster_size=15 Found 68 clusters in 0.27051661399991644s
Max Fitting with min_cluster_size=30 Found 30 clusters in 0.10185916199998246s
Max Fitting with min_cluster_size=60 Found 16 clusters in 0.10068853400025546s
Max Fitting with min_cluster_size=120 Found 10 clusters in 0.09500915999979043s
Max Fitting with min_cluster_size=240 Found 6 clusters in 0.08926718299971981s
Midpoint Fitting with min_cluster_size=180 Found 9 clusters in 0.09352127000011023s
Midpoint Fitting with min_cluster_size=210 Found 6 clusters in 0.09090468000022156s
Midpoint Fitting with min_cluster_size=195 Found 8 clusters in 0.09274479600026098s
No need Re-Fitting with min_cluster_size=195
Clusters cached: [6, 6, 8, 9, 10, 16, 30, 68]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 68 clusters in 0.10702741499972035s
Max Fitting with min_cluster_size=30 Found 30 clusters in 0.1014529180001773s
Max Fitting with min_cluster_size=60 Found 16 clusters in 0.10032602599994789s
Midpoint Fitting with min_cluster_size=45 Found 21 clusters in 0.09942904999979874s
Midpoint Fitting with min_cluster_size=37 Found 23 clusters in 0.09883285599971714s
Re-Fitting with min_cluster_size=30 Found 30 clusters in 0.10167597100007697s
Clusters cached: [16, 21, 23, 30, 68]
Cluster 11 does not have enough number of labels!
Number of labels in cluster 1.
Warning: Less than 2 words in cluster 11 with (1) words.
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : 0.3524
4_clus__avg_word_couv_0 : 0.4317
4_clus__med_word_couv_0 : 0.4484
4_clus__avg_word_couv_minus_0 : 0.4105
4_clus__big_small_ratio_0 : 4.5149
4_clus__stab_clus_0 : 0.4875
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : 0.2507
4_clus__avg_word_couv_1 : 0.5749
4_clus__med_word_couv_1 : 0.5336
4_clus__avg_word_couv_minus_1 : 0.5567
4_clus__big_small_ratio_1 : 35.3667
4_clus__stab_clus_1 : 0.1667
4_clus__avg_stab_avg : 0.3271
4_clus__avg_couv_avg : 0.5033
4_clus__clu_score : 0.4152
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 visualizations, human interaction : s 912 stb ...
1 optimization control, stochastic : s 738 stb 0...
2 cluster computing, networking internet archite...
3 logic science, verification : s 340 stb 0.50 +...
4 query, ontology : s 338 stb 0.50 + 0.57 - 0.03
5 discrete mathematics, combinatorics : s 266 st...
6 quantitative methods, bioinformatics : s 209 s...
7 natural language, french : s 202 stb 0.30 + 0....
dtype: object
VALUE : 0.4317
4_clus__clus_eval_pos_1_det
0 programs, specification : s 340 stb 0.00 + 0.4...
1 information visualization, cognitive science :...
2 display, reality : s 188 stb 0.40 + 0.48 - 0.02
3 secondary structure, protein : s 174 stb 0.00 ...
4 natural language processing, computation langu...
5 architectures, parallelism : s 142 stb 0.80 + ...
6 vertices, minimum : s 134 stb 0.30 + 0.58 - 0.03
7 fluid, numerical simulations : s 133 stb 0.80 ...
8 internet architecture, cloud radio access netw...
9 ontologies, linked : s 121 stb 0.00 + 0.50 - 0.03
10 queries, updates : s 117 stb 0.00 + 0.74 - 0.03
11 evolution strategies, multi objective : s 113 ...
12 biological, metabolic : s 110 stb 0.50 + 0.60 ...
13 covariance matrix adaptation, benchmarking : s...
14 fault, cloud computing : s 99 stb 0.00 + 0.49 ...
15 music, designers : s 94 stb 0.00 + 0.48 - 0.01
16 mixed integer linear, numerical results : s 91...
17 discrete, permutations : s 76 stb 0.00 + 0.49 ...
18 automl, training : s 69 stb 0.00 + 0.55 - 0.03
19 stabilizing, population protocols : s 67 stb 0...
20 belief propagation, condensed matter : s 66 st...
21 discrete event systems, causal : s 60 stb 0.20...
22 social sciences, social network : s 56 stb 0.0...
23 robotics, electric power : s 52 stb 0.40 + 0.6...
24 linear systems, factorization : s 43 stb 0.00 ...
25 monte carlo search, games : s 41 stb 0.00 + 0....
26 french language, signing : s 35 stb 0.30 + 0.6...
27 multi armed, bandit : s 34 stb 0.00 + 0.94 - 0.02
28 information retrieval, wikipedia : s 31 stb 0....
29 planning, matrix factorization : s 30 stb 0.00...
dtype: object
VALUE : 0.5749
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.5007
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.54
2 4_clus__clu_score : 0.42
dtype: object
VALUE : 0.5007
--------- Raw Scores ---------
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
Name: 0, dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
Name: 0, dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 10, '3_2D__rmin_dist': 0.25, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.5189
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.0320
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0446
3_2D__trustworthiness_sklearn : 0.9385
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Céline Gicquel : 0.81
1 Loïc Paulevé : 0.77
2 Sébastien Tixeuil : 0.74
3 Paola Tubaro : 0.73
4 Isabelle Guyon : 0.71
5 Nikolaus Hansen : 0.71
6 Franck Cappello : 0.66
7 Cyril Furtlehner : 0.66
8 Yann Ponty : 0.63
9 Dimo Brockhoff : 0.61
10 Sarah Cohen-Boulakia : 0.61
11 Marc Baboulin : 0.61
12 Philippe Caillou : 0.61
13 Albert Cohen : 0.61
14 Nicolas Bredeche : 0.61
15 Fatiha Zaidi : 0.59
16 Olivier Teytaud : 0.59
17 Ioana Manolescu : 0.59
18 Marc Schoenauer : 0.59
19 Sylvie Boldo : 0.58
20 Raymond Ros : 0.57
21 Tobias Isenberg : 0.57
22 Nathalie Pernelle : 0.57
23 Anne Auger : 0.56
24 Nathann Cohen : 0.54
25 Fatiha Saïs : 0.53
26 Guillaume Charpiat : 0.52
27 Jean-Daniel Fekete : 0.48
28 Petra Isenberg : 0.48
29 Claude Marché : 0.48
30 François Goasdoué : 0.47
31 Michèle Sebag : 0.46
32 Lonni Besançon : 0.45
33 Evelyne Lutton : 0.45
34 Chantal Reynaud : 0.43
35 Guillaume Melquiond : 0.42
36 Caroline Appert : 0.38
37 Alain Denise : 0.37
38 Anastasia Bezerianos : 0.36
39 Steven Martin : 0.35
40 Balázs Kégl : 0.34
41 Olivier Chapuis : 0.34
42 Michel Beaudouin-Lafon : 0.33
43 Pierre Dragicevic : 0.31
44 Emmanuel Pietriga : 0.31
45 Wendy Mackay : 0.28
46 Wendy E. Mackay : 0.27
47 Johanne Cohen : 0.25
dtype: object
VALUE : 0.5189
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 Guillaume Melquiond : -0.15
1 François Goasdoué : -0.15
2 Dimo Brockhoff : -0.14
3 Anne Auger : -0.12
4 Jean-Daniel Fekete : -0.12
5 Petra Isenberg : -0.11
6 Tobias Isenberg : -0.10
7 Fatiha Saïs : -0.09
8 Evelyne Lutton : -0.09
9 Chantal Reynaud : -0.08
10 Michel Beaudouin-Lafon : -0.08
11 Pierre Dragicevic : -0.08
12 Johanne Cohen : -0.07
13 Lonni Besançon : -0.07
14 Ioana Manolescu : -0.07
15 Steven Martin : -0.06
16 Wendy Mackay : -0.06
17 Olivier Chapuis : -0.06
18 Balázs Kégl : -0.06
19 Nikolaus Hansen : -0.06
20 Emmanuel Pietriga : -0.05
21 Céline Gicquel : -0.05
22 Claude Marché : -0.05
23 Nathalie Pernelle : -0.05
24 Marc Baboulin : -0.04
25 Guillaume Charpiat : -0.04
26 Yann Ponty : -0.03
27 Nicolas Bredeche : -0.03
28 Loïc Paulevé : -0.02
29 Philippe Caillou : -0.01
30 Franck Cappello : 0.01
31 Alain Denise : 0.01
32 Cyril Furtlehner : 0.01
33 Olivier Teytaud : 0.02
34 Nathann Cohen : 0.02
35 Anastasia Bezerianos : 0.02
36 Michèle Sebag : 0.02
37 Sylvie Boldo : 0.03
38 Marc Schoenauer : 0.03
39 Caroline Appert : 0.03
40 Wendy E. Mackay : 0.04
41 Raymond Ros : 0.04
42 Albert Cohen : 0.04
43 Fatiha Zaidi : 0.04
44 Sarah Cohen-Boulakia : 0.05
45 Paola Tubaro : 0.06
46 Isabelle Guyon : 0.06
47 Sébastien Tixeuil : 0.11
dtype: object
VALUE : -0.0320
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.736170
Michèle Sebag 0.457664
Johanne Cohen 0.253488
Albert Cohen 0.606897
Wendy E. Mackay 0.267391
Philippe Caillou 0.606977
Alain Denise 0.369444
Jean-Daniel Fekete 0.483648
Emmanuel Pietriga 0.309524
Yann Ponty 0.633333
Marc Schoenauer 0.588489
Franck Cappello 0.660976
Caroline Appert 0.376087
Michel Beaudouin-Lafon 0.332099
Wendy Mackay 0.282979
Anne Auger 0.556962
Evelyne Lutton 0.454054
Pierre Dragicevic 0.313580
Ioana Manolescu 0.592683
Nikolaus Hansen 0.706173
Nicolas Bredeche 0.605882
Olivier Teytaud 0.593396
François Goasdoué 0.473585
Nathalie Pernelle 0.570588
Fatiha Saïs 0.529268
Sarah Cohen-Boulakia 0.612121
Claude Marché 0.478723
Chantal Reynaud 0.426667
Olivier Chapuis 0.336538
Steven Martin 0.353846
Fatiha Zaidi 0.593750
Balázs Kégl 0.342105
Paola Tubaro 0.728205
Raymond Ros 0.573529
Cyril Furtlehner 0.656410
Anastasia Bezerianos 0.359701
Sylvie Boldo 0.582857
Guillaume Melquiond 0.415152
Marc Baboulin 0.608889
Dimo Brockhoff 0.612821
Nathann Cohen 0.539024
Petra Isenberg 0.482243
Tobias Isenberg 0.573504
Loïc Paulevé 0.769048
Céline Gicquel 0.807895
Isabelle Guyon 0.714607
Guillaume Charpiat 0.522581
Lonni Besançon 0.454545
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.736170 0.112766
Michèle Sebag 0.457664 0.024818
Johanne Cohen 0.253488 -0.072093
Albert Cohen 0.606897 0.041379
Wendy E. Mackay 0.267391 0.036957
Philippe Caillou 0.606977 -0.011628
Alain Denise 0.369444 0.011111
Jean-Daniel Fekete 0.483648 -0.117610
Emmanuel Pietriga 0.309524 -0.053968
Yann Ponty 0.633333 -0.030303
Marc Schoenauer 0.588489 0.029496
Franck Cappello 0.660976 0.009756
Caroline Appert 0.376087 0.034783
Michel Beaudouin-Lafon 0.332099 -0.081481
Wendy Mackay 0.282979 -0.061702
Anne Auger 0.556962 -0.120253
Evelyne Lutton 0.454054 -0.086486
Pierre Dragicevic 0.313580 -0.079012
Ioana Manolescu 0.592683 -0.067073
Nikolaus Hansen 0.706173 -0.055556
Nicolas Bredeche 0.605882 -0.029412
Olivier Teytaud 0.593396 0.017925
François Goasdoué 0.473585 -0.147170
Nathalie Pernelle 0.570588 -0.047059
Fatiha Saïs 0.529268 -0.092683
Sarah Cohen-Boulakia 0.612121 0.048485
Claude Marché 0.478723 -0.048936
Chantal Reynaud 0.426667 -0.083333
Olivier Chapuis 0.336538 -0.061538
Steven Martin 0.353846 -0.064103
Fatiha Zaidi 0.593750 0.043750
Balázs Kégl 0.342105 -0.060526
Paola Tubaro 0.728205 0.056410
Raymond Ros 0.573529 0.038235
Cyril Furtlehner 0.656410 0.012821
Anastasia Bezerianos 0.359701 0.023881
Sylvie Boldo 0.582857 0.025714
Guillaume Melquiond 0.415152 -0.148485
Marc Baboulin 0.608889 -0.042222
Dimo Brockhoff 0.612821 -0.138462
Nathann Cohen 0.539024 0.019512
Petra Isenberg 0.482243 -0.110280
Tobias Isenberg 0.573504 -0.098291
Loïc Paulevé 0.769048 -0.016667
Céline Gicquel 0.807895 -0.050000
Isabelle Guyon 0.714607 0.058427
Guillaume Charpiat 0.522581 -0.035484
Lonni Besançon 0.454545 -0.069697
Nothing in cache, initial Fitting with min_cluster_size=15 Found 56 clusters in 0.28196315100012725s
Max Fitting with min_cluster_size=30 Found 31 clusters in 0.10167976899992937s
Max Fitting with min_cluster_size=60 Found 19 clusters in 0.0976447210000515s
Max Fitting with min_cluster_size=120 Found 10 clusters in 0.09320349900008296s
Max Fitting with min_cluster_size=240 Found 5 clusters in 0.09002576800003226s
Midpoint Fitting with min_cluster_size=180 Found 5 clusters in 0.08962468700019599s
Midpoint Fitting with min_cluster_size=150 Found 7 clusters in 0.09160015399993426s
Midpoint Fitting with min_cluster_size=135 Found 8 clusters in 0.09431141200002457s
No need Re-Fitting with min_cluster_size=135
Clusters cached: [5, 5, 7, 8, 10, 19, 31, 56]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 56 clusters in 0.10585523600002489s
Max Fitting with min_cluster_size=30 Found 31 clusters in 0.10178884100014329s
Max Fitting with min_cluster_size=60 Found 19 clusters in 0.09873409099964192s
Midpoint Fitting with min_cluster_size=45 Found 21 clusters in 0.09887258800017662s
Midpoint Fitting with min_cluster_size=37 Found 23 clusters in 0.09989442599999165s
Re-Fitting with min_cluster_size=30 Found 31 clusters in 0.10044263299960221s
Clusters cached: [19, 21, 23, 31, 56]
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : 0.1678
4_clus__avg_word_couv_0 : 0.4404
4_clus__med_word_couv_0 : 0.4362
4_clus__avg_word_couv_minus_0 : 0.4153
4_clus__big_small_ratio_0 : 9.2214
4_clus__stab_clus_0 : 0.1000
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : 0.1259
4_clus__avg_word_couv_1 : 0.5946
4_clus__med_word_couv_1 : 0.6053
4_clus__avg_word_couv_minus_1 : 0.5780
4_clus__big_small_ratio_1 : 37.9355
4_clus__stab_clus_1 : 0.0968
4_clus__avg_stab_avg : 0.0984
4_clus__avg_couv_avg : 0.5175
4_clus__clu_score : 0.3079
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 visualizations, interaction techniques : s 743...
1 quantitative methods, discrete mathematics : s...
2 query, cluster computing : s 496 stb 0.00 + 0....
3 optimization control, covariance matrix adapta...
4 logic science, verification : s 278 stb 0.20 +...
5 networking internet architecture, protocols : ...
6 natural language, computation language : s 152...
7 architectures, parallelism : s 140 stb 0.00 + ...
dtype: object
VALUE : 0.4404
4_clus__clus_eval_pos_1_det
0 information visualization, interfaces : s 695 ...
1 specification, programs : s 278 stb 0.00 + 0.4...
2 ontology, linked : s 157 stb 0.40 + 0.49 - 0.03
3 natural language processing, speech : s 152 st...
4 compiler, hardware : s 140 stb 0.00 + 0.54 - 0.03
5 combinatorics, discrete : s 125 stb 0.00 + 0.4...
6 vertices, minimum : s 117 stb 0.00 + 0.61 - 0.03
7 fluid, numerical simulations : s 101 stb 0.30 ...
8 training, automl : s 93 stb 0.00 + 0.41 - 0.03
9 biological, metabolism : s 89 stb 0.00 + 0.63 ...
10 covariance matrix adaptation evolution, functi...
11 secondary structure, genes : s 87 stb 0.00 + 0...
12 fault, cloud computing : s 80 stb 0.00 + 0.57 ...
13 stabilizing, population protocols : s 77 stb 0...
14 linear systems, factorization : s 72 stb 0.70 ...
15 internet architecture, routing : s 72 stb 0.00...
16 monte carlo search, multi armed : s 65 stb 0.0...
17 floating point, arithmetic : s 62 stb 0.90 + 0...
18 radio, resource allocation : s 58 stb 0.00 + 0...
19 queries, query answering : s 54 stb 0.00 + 0.7...
20 scientific workflows, personality : s 52 stb 0...
21 evolution strategies, noisy optimization : s 4...
22 evolutionary computation, genetic programming ...
23 belief propagation, condensed matter : s 42 st...
24 integer linear, chance constrained : s 38 stb ...
25 recommender systems, multi agent : s 36 stb 0....
26 french language, signing : s 36 stb 0.00 + 0.6...
27 workers, social network : s 34 stb 0.00 + 0.47...
28 materialized views, engines : s 33 stb 0.00 + ...
29 evolutionary robotics, autonomous : s 33 stb 0...
30 mobile robots, cartesian : s 31 stb 0.40 + 0.8...
dtype: object
VALUE : 0.5946
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.4592
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.52
2 4_clus__clu_score : 0.31
dtype: object
VALUE : 0.4592
--------- Raw Scores ---------
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
Name: 0, dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
Name: 0, dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 10, '3_2D__rmin_dist': 0.5, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.4772
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.0737
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0804
3_2D__trustworthiness_sklearn : 0.9084
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Loïc Paulevé : 0.76
1 Céline Gicquel : 0.74
2 Isabelle Guyon : 0.69
3 Sébastien Tixeuil : 0.69
4 Sarah Cohen-Boulakia : 0.66
5 Marc Baboulin : 0.66
6 Nikolaus Hansen : 0.64
7 Paola Tubaro : 0.63
8 Nicolas Bredeche : 0.62
9 Cyril Furtlehner : 0.62
10 Franck Cappello : 0.57
11 Dimo Brockhoff : 0.57
12 Olivier Teytaud : 0.56
13 Raymond Ros : 0.56
14 Ioana Manolescu : 0.56
15 Albert Cohen : 0.56
16 Yann Ponty : 0.55
17 Sylvie Boldo : 0.53
18 Anne Auger : 0.53
19 Philippe Caillou : 0.50
20 Jean-Daniel Fekete : 0.49
21 Marc Schoenauer : 0.49
22 Nathann Cohen : 0.47
23 Guillaume Charpiat : 0.47
24 Tobias Isenberg : 0.47
25 François Goasdoué : 0.47
26 Claude Marché : 0.45
27 Guillaume Melquiond : 0.45
28 Fatiha Saïs : 0.43
29 Lonni Besançon : 0.43
30 Petra Isenberg : 0.42
31 Chantal Reynaud : 0.42
32 Evelyne Lutton : 0.42
33 Fatiha Zaidi : 0.41
34 Steven Martin : 0.39
35 Michèle Sebag : 0.38
36 Nathalie Pernelle : 0.38
37 Olivier Chapuis : 0.35
38 Alain Denise : 0.35
39 Caroline Appert : 0.34
40 Balázs Kégl : 0.31
41 Anastasia Bezerianos : 0.31
42 Pierre Dragicevic : 0.31
43 Michel Beaudouin-Lafon : 0.30
44 Johanne Cohen : 0.27
45 Wendy Mackay : 0.26
46 Wendy E. Mackay : 0.25
47 Emmanuel Pietriga : 0.24
dtype: object
VALUE : 0.4772
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 Nathalie Pernelle : -0.24
1 Tobias Isenberg : -0.20
2 Fatiha Saïs : -0.19
3 Dimo Brockhoff : -0.18
4 Petra Isenberg : -0.17
5 François Goasdoué : -0.15
6 Anne Auger : -0.15
7 Fatiha Zaidi : -0.14
8 Evelyne Lutton : -0.12
9 Céline Gicquel : -0.12
10 Emmanuel Pietriga : -0.12
11 Philippe Caillou : -0.12
12 Nikolaus Hansen : -0.12
13 Guillaume Melquiond : -0.12
14 Yann Ponty : -0.12
15 Michel Beaudouin-Lafon : -0.11
16 Jean-Daniel Fekete : -0.11
17 Ioana Manolescu : -0.10
18 Lonni Besançon : -0.09
19 Chantal Reynaud : -0.09
20 Balázs Kégl : -0.09
21 Guillaume Charpiat : -0.09
22 Wendy Mackay : -0.09
23 Pierre Dragicevic : -0.08
24 Franck Cappello : -0.08
25 Claude Marché : -0.08
26 Marc Schoenauer : -0.07
27 Johanne Cohen : -0.06
28 Michèle Sebag : -0.05
29 Olivier Chapuis : -0.05
30 Nathann Cohen : -0.05
31 Paola Tubaro : -0.05
32 Steven Martin : -0.03
33 Cyril Furtlehner : -0.03
34 Sylvie Boldo : -0.03
35 Anastasia Bezerianos : -0.02
36 Loïc Paulevé : -0.02
37 Nicolas Bredeche : -0.02
38 Olivier Teytaud : -0.01
39 Alain Denise : -0.01
40 Albert Cohen : -0.01
41 Caroline Appert : -0.00
42 Marc Baboulin : 0.01
43 Wendy E. Mackay : 0.02
44 Raymond Ros : 0.03
45 Isabelle Guyon : 0.03
46 Sébastien Tixeuil : 0.06
47 Sarah Cohen-Boulakia : 0.10
dtype: object
VALUE : -0.0737
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.685106
Michèle Sebag 0.383942
Johanne Cohen 0.269767
Albert Cohen 0.556897
Wendy E. Mackay 0.245652
Philippe Caillou 0.500000
Alain Denise 0.347222
Jean-Daniel Fekete 0.493711
Emmanuel Pietriga 0.244444
Yann Ponty 0.546970
Marc Schoenauer 0.493525
Franck Cappello 0.573171
Caroline Appert 0.339130
Michel Beaudouin-Lafon 0.298765
Wendy Mackay 0.259574
Anne Auger 0.525316
Evelyne Lutton 0.416216
Pierre Dragicevic 0.309877
Ioana Manolescu 0.560976
Nikolaus Hansen 0.643210
Nicolas Bredeche 0.617647
Olivier Teytaud 0.563208
François Goasdoué 0.466038
Nathalie Pernelle 0.376471
Fatiha Saïs 0.431707
Sarah Cohen-Boulakia 0.663636
Claude Marché 0.451064
Chantal Reynaud 0.416667
Olivier Chapuis 0.350000
Steven Martin 0.387179
Fatiha Zaidi 0.406250
Balázs Kégl 0.313158
Paola Tubaro 0.625641
Raymond Ros 0.561765
Cyril Furtlehner 0.615385
Anastasia Bezerianos 0.311940
Sylvie Boldo 0.531429
Guillaume Melquiond 0.445455
Marc Baboulin 0.662222
Dimo Brockhoff 0.569231
Nathann Cohen 0.473171
Petra Isenberg 0.417757
Tobias Isenberg 0.467521
Loïc Paulevé 0.761905
Céline Gicquel 0.736842
Isabelle Guyon 0.687640
Guillaume Charpiat 0.470968
Lonni Besançon 0.430303
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.685106 0.061702
Michèle Sebag 0.383942 -0.048905
Johanne Cohen 0.269767 -0.055814
Albert Cohen 0.556897 -0.008621
Wendy E. Mackay 0.245652 0.015217
Philippe Caillou 0.500000 -0.118605
Alain Denise 0.347222 -0.011111
Jean-Daniel Fekete 0.493711 -0.107547
Emmanuel Pietriga 0.244444 -0.119048
Yann Ponty 0.546970 -0.116667
Marc Schoenauer 0.493525 -0.065468
Franck Cappello 0.573171 -0.078049
Caroline Appert 0.339130 -0.002174
Michel Beaudouin-Lafon 0.298765 -0.114815
Wendy Mackay 0.259574 -0.085106
Anne Auger 0.525316 -0.151899
Evelyne Lutton 0.416216 -0.124324
Pierre Dragicevic 0.309877 -0.082716
Ioana Manolescu 0.560976 -0.098780
Nikolaus Hansen 0.643210 -0.118519
Nicolas Bredeche 0.617647 -0.017647
Olivier Teytaud 0.563208 -0.012264
François Goasdoué 0.466038 -0.154717
Nathalie Pernelle 0.376471 -0.241176
Fatiha Saïs 0.431707 -0.190244
Sarah Cohen-Boulakia 0.663636 0.100000
Claude Marché 0.451064 -0.076596
Chantal Reynaud 0.416667 -0.093333
Olivier Chapuis 0.350000 -0.048077
Steven Martin 0.387179 -0.030769
Fatiha Zaidi 0.406250 -0.143750
Balázs Kégl 0.313158 -0.089474
Paola Tubaro 0.625641 -0.046154
Raymond Ros 0.561765 0.026471
Cyril Furtlehner 0.615385 -0.028205
Anastasia Bezerianos 0.311940 -0.023881
Sylvie Boldo 0.531429 -0.025714
Guillaume Melquiond 0.445455 -0.118182
Marc Baboulin 0.662222 0.011111
Dimo Brockhoff 0.569231 -0.182051
Nathann Cohen 0.473171 -0.046341
Petra Isenberg 0.417757 -0.174766
Tobias Isenberg 0.467521 -0.204274
Loïc Paulevé 0.761905 -0.023810
Céline Gicquel 0.736842 -0.121053
Isabelle Guyon 0.687640 0.031461
Guillaume Charpiat 0.470968 -0.087097
Lonni Besançon 0.430303 -0.093939
Nothing in cache, initial Fitting with min_cluster_size=15 Found 7 clusters in 0.3022271129998444s
No need Re-Fitting with min_cluster_size=15
Clusters cached: [7]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 7 clusters in 0.11288243899980444s
No need Re-Fitting with min_cluster_size=15
Clusters cached: [7]
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : -0.2171
4_clus__avg_word_couv_0 : 0.6430
4_clus__med_word_couv_0 : 0.6818
4_clus__avg_word_couv_minus_0 : 0.6382
4_clus__big_small_ratio_0 : 252.5625
4_clus__stab_clus_0 : 0.0000
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : -0.2171
4_clus__avg_word_couv_1 : 0.6226
4_clus__med_word_couv_1 : 0.6818
4_clus__avg_word_couv_minus_1 : 0.6161
4_clus__big_small_ratio_1 : 252.5625
4_clus__stab_clus_1 : 0.0000
4_clus__avg_stab_avg : 0.0000
4_clus__avg_couv_avg : 0.6328
4_clus__clu_score : 0.3164
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 query, verification : s 4041 stb 0.00 + 0.09 -...
1 large hadron collider, astrophysics : s 44 stb...
2 french language, signing : s 35 stb 0.00 + 0.6...
3 evolutionary robotics, autonomous : s 33 stb 0...
4 mobile robots : s 18 stb 0.00 + 0.83 - 0.00
5 approximate bayesian, genetics : s 17 stb 0.00...
6 electric power, power grids : s 16 stb 0.00 + ...
dtype: object
VALUE : 0.6430
4_clus__clus_eval_pos_1_det
0 memory, quantitative methods : s 4041 stb 0.00...
1 energy physics, cosmic : s 44 stb 0.00 + 0.68 ...
2 motion capture, signs : s 35 stb 0.00 + 0.51 -...
3 robotics : s 33 stb 0.00 + 0.76 - 0.01
4 robots : s 18 stb 0.00 + 0.94 - 0.01
5 demographic, populations : s 17 stb 0.00 + 0.8...
6 network architecture, electricity : s 16 stb 0...
dtype: object
VALUE : 0.6226
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.4482
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.48
2 4_clus__clu_score : 0.32
dtype: object
VALUE : 0.4482
--------- Raw Scores ---------
.. GENERATED FROM PYTHON SOURCE LINES 128-129
When the experiment is run, results of all runs is saved in `experiment.results`. We access the values corresponding to each run with `experiment.results.runs_`.
.. GENERATED FROM PYTHON SOURCE LINES 129-147
.. code-block:: Python
experiment.results.runs_[0].scores
""
list(experiment.results.runs_[0].desc_scores.keys())
""
experiment.results.runs_[0].desc_scores
""
list(experiment.results.runs_[0].raw_scores.keys())
""
experiment.results.runs_[0].raw_scores
""
experiment.results.print_best(n=20)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
------------------------
Best 20 results:
------------------------
0_agscore id 2_nD__rkey 2_nD__rnum_dims \
0 0.500707 5acf1c353d2c392a1729 bert 768
1 0.459228 da34d4237bc01ee70684 bert 768
2 0.448154 a862f98f80e4912e4e08 bert 768
2_nD__rnormalize 2_nD__rfamily 2_nD__rmax_length 3_2D__rkey \
0 True all-MiniLM-L6-v2 256 umap
1 True all-MiniLM-L6-v2 256 umap
2 True all-MiniLM-L6-v2 256 umap
3_2D__rmetric 3_2D__rn_neighbors 3_2D__rmin_dist 3_2D__rinit \
0 euclidean 10 0.10 random
1 euclidean 10 0.25 random
2 euclidean 10 0.50 random
3_2D__rlearning_rate 3_2D__rn_epochs 3_2D__rrandom_state 4_clus__rkey \
0 1.0 None None hdbscan
1 1.0 None None hdbscan
2 1.0 None None hdbscan
4_clus__rbase_factor 2_nD__smin_score 2_nD__srecompute \
0 3 30 True
1 3 30 True
2 3 30 True
2_nD__ssample_size 2_nD__sn_neighbors 2_nD__srandom_state \
0 None 10 42
1 None 10 42
2 None 10 42
3_2D__smin_score 3_2D__srecompute 3_2D__ssample_size 3_2D__sn_neighbors \
0 30 True None 10
1 30 True None 10
2 30 True None 10
3_2D__srandom_state 3_2D__smetric 4_clus__siter_stab \
0 42 euclidean 2
1 42 euclidean 2
2 42 euclidean 2
4_clus__sremove_stab 4_clus__smetric 4_clus__srandom_state \
0 [0, 0.01, 0.03, 0.1, 0.25] euclidean None
1 [0, 0.01, 0.03, 0.1, 0.25] euclidean None
2 [0, 0.01, 0.03, 0.1, 0.25] euclidean None
6_pst__sname_list \
0 [2_nD__neighbors_articles_authors, 3_2D__neigh...
1 [2_nD__neighbors_articles_authors, 3_2D__neigh...
2 [2_nD__neighbors_articles_authors, 3_2D__neigh...
2_nD__neighbors_articles_authors 3_2D__neighbors_articles_authors \
0 0.550862 0.536061
1 0.550862 0.518877
2 0.550862 0.477201
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean \
0 -0.014801
1 -0.031985
2 -0.073661
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median \
0 -0.011051
1 -0.044641
2 -0.080382
3_2D__trustworthiness_sklearn 4_clus__nb_clust_0 4_clus__silhouette_0 \
0 0.948647 8 0.352386
1 0.938497 8 0.167785
2 0.908424 8 -0.217119
4_clus__avg_word_couv_0 4_clus__med_word_couv_0 \
0 0.431710 0.448429
1 0.440386 0.436227
2 0.643001 0.681818
4_clus__avg_word_couv_minus_0 4_clus__big_small_ratio_0 \
0 0.410497 4.514851
1 0.415253 9.221429
2 0.638169 252.562500
4_clus__stab_clus_0 4_clus__nb_clust_1 4_clus__silhouette_1 \
0 0.4875 24 0.250664
1 0.1000 24 0.125918
2 0.0000 24 -0.217119
4_clus__avg_word_couv_1 4_clus__med_word_couv_1 \
0 0.574919 0.533592
1 0.594620 0.605263
2 0.622592 0.681818
4_clus__avg_word_couv_minus_1 4_clus__big_small_ratio_1 \
0 0.556723 35.366667
1 0.578013 37.935484
2 0.616136 252.562500
4_clus__stab_clus_1 4_clus__avg_stab_avg 4_clus__avg_couv_avg \
0 0.166667 0.327083 0.503315
1 0.096774 0.098387 0.517503
2 0.000000 0.000000 0.632796
4_clus__clu_score 6_pst__final_score \
0 0.415199 0.500707
1 0.307945 0.459228
2 0.316398 0.448154
dump active
0 experiment_pipeline_lisn/lisn/2022.11.15.1/0/m... True
1 experiment_pipeline_lisn/lisn/2022.11.15.1/0/m... True
2 experiment_pipeline_lisn/lisn/2022.11.15.1/0/m... True
.. GENERATED FROM PYTHON SOURCE LINES 148-151
Let's see some of the results of the experiment from the file system.
We will first check the contents of the `scores` directory.
.. GENERATED FROM PYTHON SOURCE LINES 151-154
.. code-block:: Python
# !ls $TOP_DIR/scores
.. GENERATED FROM PYTHON SOURCE LINES 155-156
`6_pst__final_scores.csv` file contains the final scores for each set of parameters together with the parameter values. We had 3 runs, so there are 3 values.
.. GENERATED FROM PYTHON SOURCE LINES 156-159
.. code-block:: Python
# !cat $TOP_DIR/scores/6_pst__final_score.csv
.. GENERATED FROM PYTHON SOURCE LINES 160-161
The `final_results.csv` displays each score calculated during the experiment in separate columns together with `rank` and an aggregated score `agscore`. This value is the same as the value in `6_post__final_score.csv` file.
.. GENERATED FROM PYTHON SOURCE LINES 161-164
.. code-block:: Python
# !cat $TOP_DIR/scores/final_results.csv
.. GENERATED FROM PYTHON SOURCE LINES 165-166
The other files in the directory contains the single score calculated for all runs. For example `2_nD__neighbors_articles_authors.csv` file contains the `2_nD__neighbors_articles_authors` scores for 3 runs.
.. GENERATED FROM PYTHON SOURCE LINES 166-169
.. code-block:: Python
# !cat $TOP_DIR/scores/2_nD__neighbors_articles_authors.csv
.. GENERATED FROM PYTHON SOURCE LINES 170-171
The files that contain `det` in theirs names, contains the neighbors and their scores used to calculate the score `2_nD__neighbors_articles_authors` for each run.
.. GENERATED FROM PYTHON SOURCE LINES 171-174
.. code-block:: Python
# !cat $TOP_DIR/scores/2_nD__neighbors_articles_authors_det.csv
.. GENERATED FROM PYTHON SOURCE LINES 175-178
These files also reside in the hierarchical dataset directories generated during the run.
For example `experiment_pipeline/lisn/2022.11.15.1/0/mat_articles__authors_4_teams_4_labs_4_words_10_0.05_None_None_5_4/bert_768_True_all-MiniLM-L6-v2_256/scores_2_nD__neighbors_articles_authors_det.csv` file, but only for the specific run together with hyperparameters and scorşng parameters.
.. GENERATED FROM PYTHON SOURCE LINES 178-181
.. code-block:: Python
# !cat $TOP_DIR/lisn/2022.11.15.1/0/mat_articles__authors_4_teams_4_labs_4_words_10_0.05_None_None_5_4/bert_768_True_all-MiniLM-L6-v2_256/scores_2_nD__neighbors_articles_authors_det.csv
.. GENERATED FROM PYTHON SOURCE LINES 182-187
Actually for each set of parameters, the estimator generates a directory structure of the form:
`top_dir / dataset / dataset_version / robustseed / dataset_column_parameters / projection_nd_key_dim / projection2D_key_n_neighbors_min_dist_metric_init_learning_rate_repulsion_strength / clustering_key_base_factor`.
Each score calculated at a certain level in the directory structure is saved in that directory.
.. GENERATED FROM PYTHON SOURCE LINES 189-190
Now, we will continue the experiment to run for 3 more set of parameters.
.. GENERATED FROM PYTHON SOURCE LINES 190-193
.. code-block:: Python
experiment.run(3)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
Name: 0, dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
Name: 0, dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 20, '3_2D__rmin_dist': 0.1, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.5220
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.0289
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0334
3_2D__trustworthiness_sklearn : 0.9410
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Loïc Paulevé : 0.79
1 Paola Tubaro : 0.78
2 Céline Gicquel : 0.77
3 Isabelle Guyon : 0.72
4 Nikolaus Hansen : 0.72
5 Sarah Cohen-Boulakia : 0.69
6 Franck Cappello : 0.68
7 Nicolas Bredeche : 0.64
8 Yann Ponty : 0.64
9 Sébastien Tixeuil : 0.64
10 Tobias Isenberg : 0.63
11 Fatiha Saïs : 0.62
12 Dimo Brockhoff : 0.61
13 Marc Baboulin : 0.60
14 Cyril Furtlehner : 0.60
15 Ioana Manolescu : 0.59
16 Olivier Teytaud : 0.59
17 Philippe Caillou : 0.58
18 Albert Cohen : 0.57
19 Raymond Ros : 0.56
20 Marc Schoenauer : 0.56
21 Chantal Reynaud : 0.55
22 François Goasdoué : 0.55
23 Guillaume Charpiat : 0.55
24 Nathalie Pernelle : 0.54
25 Sylvie Boldo : 0.53
26 Anne Auger : 0.53
27 Jean-Daniel Fekete : 0.52
28 Nathann Cohen : 0.51
29 Lonni Besançon : 0.51
30 Petra Isenberg : 0.49
31 Fatiha Zaidi : 0.48
32 Claude Marché : 0.47
33 Evelyne Lutton : 0.45
34 Michèle Sebag : 0.44
35 Guillaume Melquiond : 0.42
36 Steven Martin : 0.38
37 Michel Beaudouin-Lafon : 0.38
38 Alain Denise : 0.36
39 Olivier Chapuis : 0.35
40 Balázs Kégl : 0.34
41 Caroline Appert : 0.34
42 Anastasia Bezerianos : 0.34
43 Wendy Mackay : 0.31
44 Pierre Dragicevic : 0.31
45 Emmanuel Pietriga : 0.30
46 Wendy E. Mackay : 0.26
47 Johanne Cohen : 0.25
dtype: object
VALUE : 0.5220
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 Guillaume Melquiond : -0.15
1 Anne Auger : -0.15
2 Dimo Brockhoff : -0.14
3 Petra Isenberg : -0.10
4 Evelyne Lutton : -0.09
5 Céline Gicquel : -0.09
6 Pierre Dragicevic : -0.09
7 Jean-Daniel Fekete : -0.08
8 Nathalie Pernelle : -0.07
9 Johanne Cohen : -0.07
10 Ioana Manolescu : -0.07
11 François Goasdoué : -0.07
12 Fatiha Zaidi : -0.07
13 Emmanuel Pietriga : -0.06
14 Balázs Kégl : -0.06
15 Claude Marché : -0.06
16 Marc Baboulin : -0.05
17 Nikolaus Hansen : -0.05
18 Olivier Chapuis : -0.04
19 Cyril Furtlehner : -0.04
20 Tobias Isenberg : -0.04
21 Michel Beaudouin-Lafon : -0.04
22 Steven Martin : -0.04
23 Philippe Caillou : -0.03
24 Wendy Mackay : -0.03
25 Sylvie Boldo : -0.02
26 Yann Ponty : -0.02
27 Lonni Besançon : -0.02
28 Guillaume Charpiat : -0.01
29 Nathann Cohen : -0.01
30 Caroline Appert : -0.00
31 Fatiha Saïs : 0.00
32 Marc Schoenauer : 0.00
33 Alain Denise : 0.00
34 Anastasia Bezerianos : 0.00
35 Albert Cohen : 0.00
36 Loïc Paulevé : 0.00
37 Michèle Sebag : 0.01
38 Nicolas Bredeche : 0.01
39 Olivier Teytaud : 0.01
40 Sébastien Tixeuil : 0.02
41 Wendy E. Mackay : 0.03
42 Raymond Ros : 0.03
43 Franck Cappello : 0.03
44 Chantal Reynaud : 0.04
45 Isabelle Guyon : 0.06
46 Paola Tubaro : 0.11
47 Sarah Cohen-Boulakia : 0.13
dtype: object
VALUE : -0.0289
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.640426
Michèle Sebag 0.437956
Johanne Cohen 0.253488
Albert Cohen 0.568966
Wendy E. Mackay 0.256522
Philippe Caillou 0.583721
Alain Denise 0.361111
Jean-Daniel Fekete 0.522013
Emmanuel Pietriga 0.300000
Yann Ponty 0.643939
Marc Schoenauer 0.559712
Franck Cappello 0.680488
Caroline Appert 0.339130
Michel Beaudouin-Lafon 0.375309
Wendy Mackay 0.312766
Anne Auger 0.529114
Evelyne Lutton 0.445946
Pierre Dragicevic 0.306173
Ioana Manolescu 0.591463
Nikolaus Hansen 0.716049
Nicolas Bredeche 0.644118
Olivier Teytaud 0.587736
François Goasdoué 0.552830
Nathalie Pernelle 0.544118
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.690909
Claude Marché 0.465957
Chantal Reynaud 0.553333
Olivier Chapuis 0.353846
Steven Martin 0.382051
Fatiha Zaidi 0.484375
Balázs Kégl 0.339474
Paola Tubaro 0.784615
Raymond Ros 0.561765
Cyril Furtlehner 0.600000
Anastasia Bezerianos 0.338806
Sylvie Boldo 0.534286
Guillaume Melquiond 0.415152
Marc Baboulin 0.604444
Dimo Brockhoff 0.610256
Nathann Cohen 0.512195
Petra Isenberg 0.487850
Tobias Isenberg 0.629915
Loïc Paulevé 0.790476
Céline Gicquel 0.768421
Isabelle Guyon 0.719101
Guillaume Charpiat 0.545161
Lonni Besançon 0.509091
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.640426 0.017021
Michèle Sebag 0.437956 0.005109
Johanne Cohen 0.253488 -0.072093
Albert Cohen 0.568966 0.003448
Wendy E. Mackay 0.256522 0.026087
Philippe Caillou 0.583721 -0.034884
Alain Denise 0.361111 0.002778
Jean-Daniel Fekete 0.522013 -0.079245
Emmanuel Pietriga 0.300000 -0.063492
Yann Ponty 0.643939 -0.019697
Marc Schoenauer 0.559712 0.000719
Franck Cappello 0.680488 0.029268
Caroline Appert 0.339130 -0.002174
Michel Beaudouin-Lafon 0.375309 -0.038272
Wendy Mackay 0.312766 -0.031915
Anne Auger 0.529114 -0.148101
Evelyne Lutton 0.445946 -0.094595
Pierre Dragicevic 0.306173 -0.086420
Ioana Manolescu 0.591463 -0.068293
Nikolaus Hansen 0.716049 -0.045679
Nicolas Bredeche 0.644118 0.008824
Olivier Teytaud 0.587736 0.012264
François Goasdoué 0.552830 -0.067925
Nathalie Pernelle 0.544118 -0.073529
Fatiha Saïs 0.621951 0.000000
Sarah Cohen-Boulakia 0.690909 0.127273
Claude Marché 0.465957 -0.061702
Chantal Reynaud 0.553333 0.043333
Olivier Chapuis 0.353846 -0.044231
Steven Martin 0.382051 -0.035897
Fatiha Zaidi 0.484375 -0.065625
Balázs Kégl 0.339474 -0.063158
Paola Tubaro 0.784615 0.112821
Raymond Ros 0.561765 0.026471
Cyril Furtlehner 0.600000 -0.043590
Anastasia Bezerianos 0.338806 0.002985
Sylvie Boldo 0.534286 -0.022857
Guillaume Melquiond 0.415152 -0.148485
Marc Baboulin 0.604444 -0.046667
Dimo Brockhoff 0.610256 -0.141026
Nathann Cohen 0.512195 -0.007317
Petra Isenberg 0.487850 -0.104673
Tobias Isenberg 0.629915 -0.041880
Loïc Paulevé 0.790476 0.004762
Céline Gicquel 0.768421 -0.089474
Isabelle Guyon 0.719101 0.062921
Guillaume Charpiat 0.545161 -0.012903
Lonni Besançon 0.509091 -0.015152
Nothing in cache, initial Fitting with min_cluster_size=15 Found 61 clusters in 0.25722732500025813s
Max Fitting with min_cluster_size=30 Found 27 clusters in 0.1010039620000498s
Max Fitting with min_cluster_size=60 Found 15 clusters in 0.09871437099991454s
Max Fitting with min_cluster_size=120 Found 11 clusters in 0.0962613040001088s
Max Fitting with min_cluster_size=240 Found 4 clusters in 0.09473326699981044s
Midpoint Fitting with min_cluster_size=180 Found 9 clusters in 0.09300725299999613s
Midpoint Fitting with min_cluster_size=210 Found 5 clusters in 0.09343649799984632s
Midpoint Fitting with min_cluster_size=195 Found 9 clusters in 0.09285305200000948s
Midpoint Fitting with min_cluster_size=202 Found 9 clusters in 0.11084757000026002s
No need Re-Fitting with min_cluster_size=202
Clusters cached: [4, 5, 9, 9, 9, 11, 15, 27, 61]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 61 clusters in 0.10352430000011736s
Max Fitting with min_cluster_size=30 Found 27 clusters in 0.10104787399995985s
Max Fitting with min_cluster_size=60 Found 15 clusters in 0.0988867040000514s
Midpoint Fitting with min_cluster_size=45 Found 18 clusters in 0.10105969500000356s
Midpoint Fitting with min_cluster_size=37 Found 19 clusters in 0.10129343100015831s
Re-Fitting with min_cluster_size=30 Found 27 clusters in 0.10129774400002134s
Clusters cached: [15, 18, 19, 27, 61]
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : 0.3710
4_clus__avg_word_couv_0 : 0.4423
4_clus__med_word_couv_0 : 0.4916
4_clus__avg_word_couv_minus_0 : 0.4208
4_clus__big_small_ratio_0 : 3.4880
4_clus__stab_clus_0 : 0.5111
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : 0.2395
4_clus__avg_word_couv_1 : 0.6312
4_clus__med_word_couv_1 : 0.6190
4_clus__avg_word_couv_minus_1 : 0.6108
4_clus__big_small_ratio_1 : 29.1000
4_clus__stab_clus_1 : 0.0593
4_clus__avg_stab_avg : 0.2852
4_clus__avg_couv_avg : 0.5367
4_clus__clu_score : 0.4110
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 visualizations, human interaction : s 725 stb ...
1 networking internet architecture, cluster comp...
2 query, ontology : s 415 stb 0.00 + 0.49 - 0.02
3 logic science, verification : s 375 stb 0.70 +...
4 optimization control, covariance matrix adapta...
5 quantitative methods, bioinformatics : s 301 s...
6 neural networks, image : s 292 stb 0.00 + 0.35...
7 architectures, parallelism : s 225 stb 0.70 + ...
8 natural language, french : s 209 stb 0.00 + 0....
dtype: object
VALUE : 0.4423
4_clus__clus_eval_pos_1_det
0 information visualization, interaction techniq...
1 programs, specification : s 375 stb 0.00 + 0.4...
2 training, image processing : s 292 stb 0.00 + ...
3 natural language processing, computation langu...
4 ontologies, linked : s 163 stb 0.00 + 0.43 - 0.03
5 compiler, hardware : s 140 stb 0.00 + 0.56 - 0.03
6 queries, resource description framework : s 13...
7 networking internet, stabilizing : s 137 stb 0...
8 fluid, numerical simulations : s 130 stb 0.90 ...
9 vertices, minimum : s 116 stb 0.00 + 0.61 - 0.03
10 biological, metabolic : s 105 stb 0.20 + 0.62 ...
11 secondary structure, sequence : s 100 stb 0.00...
12 covariance matrix adaptation evolution, benchm...
13 combinatorics, finite : s 81 stb 0.00 + 0.52 -...
14 fault, cloud computing : s 80 stb 0.00 + 0.59 ...
15 mixed integer linear, stochastic : s 78 stb 0....
16 evolution strategies, noisy optimization : s 7...
17 radio, resource allocation : s 53 stb 0.00 + 0...
18 humanities social sciences, social networks : ...
19 linear systems, factorization : s 46 stb 0.00 ...
20 agent : s 41 stb 0.00 + 0.78 - 0.01
21 monte carlo search, games : s 40 stb 0.00 + 0....
22 french language, signing : s 34 stb 0.00 + 0.6...
23 multi armed, bandit : s 32 stb 0.00 + 0.94 - 0.02
24 protein : s 31 stb 0.00 + 0.87 - 0.01
25 scientific workflows, workflow : s 30 stb 0.00...
26 evolutionary robotics, autonomous : s 30 stb 0...
dtype: object
VALUE : 0.6312
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.4946
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.52
2 4_clus__clu_score : 0.41
dtype: object
VALUE : 0.4946
--------- Raw Scores ---------
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
Name: 0, dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
Name: 0, dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 20, '3_2D__rmin_dist': 0.25, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.5002
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.0506
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0525
3_2D__trustworthiness_sklearn : 0.9295
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Céline Gicquel : 0.82
1 Paola Tubaro : 0.76
2 Loïc Paulevé : 0.72
3 Sarah Cohen-Boulakia : 0.67
4 Nikolaus Hansen : 0.65
5 Isabelle Guyon : 0.64
6 Franck Cappello : 0.64
7 Dimo Brockhoff : 0.64
8 Sébastien Tixeuil : 0.63
9 Cyril Furtlehner : 0.61
10 Philippe Caillou : 0.60
11 Nicolas Bredeche : 0.60
12 Yann Ponty : 0.59
13 Ioana Manolescu : 0.59
14 Marc Baboulin : 0.59
15 Fatiha Saïs : 0.58
16 Tobias Isenberg : 0.58
17 Olivier Teytaud : 0.57
18 Albert Cohen : 0.56
19 Nathalie Pernelle : 0.54
20 Nathann Cohen : 0.54
21 Anne Auger : 0.54
22 Raymond Ros : 0.53
23 François Goasdoué : 0.52
24 Jean-Daniel Fekete : 0.51
25 Guillaume Charpiat : 0.51
26 Marc Schoenauer : 0.51
27 Sylvie Boldo : 0.49
28 Claude Marché : 0.49
29 Fatiha Zaidi : 0.49
30 Petra Isenberg : 0.45
31 Chantal Reynaud : 0.45
32 Lonni Besançon : 0.44
33 Guillaume Melquiond : 0.44
34 Michèle Sebag : 0.42
35 Steven Martin : 0.38
36 Michel Beaudouin-Lafon : 0.37
37 Olivier Chapuis : 0.37
38 Evelyne Lutton : 0.36
39 Pierre Dragicevic : 0.34
40 Wendy Mackay : 0.34
41 Anastasia Bezerianos : 0.32
42 Balázs Kégl : 0.32
43 Alain Denise : 0.30
44 Caroline Appert : 0.27
45 Emmanuel Pietriga : 0.27
46 Wendy E. Mackay : 0.26
47 Johanne Cohen : 0.24
dtype: object
VALUE : 0.5002
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 Evelyne Lutton : -0.18
1 Anne Auger : -0.14
2 Petra Isenberg : -0.14
3 Guillaume Melquiond : -0.13
4 Dimo Brockhoff : -0.12
5 Nikolaus Hansen : -0.11
6 François Goasdoué : -0.10
7 Emmanuel Pietriga : -0.10
8 Tobias Isenberg : -0.09
9 Balázs Kégl : -0.09
10 Jean-Daniel Fekete : -0.09
11 Lonni Besançon : -0.08
12 Johanne Cohen : -0.08
13 Caroline Appert : -0.08
14 Nathalie Pernelle : -0.07
15 Yann Ponty : -0.07
16 Ioana Manolescu : -0.07
17 Marc Baboulin : -0.06
18 Loïc Paulevé : -0.06
19 Sylvie Boldo : -0.06
20 Fatiha Zaidi : -0.06
21 Alain Denise : -0.06
22 Chantal Reynaud : -0.06
23 Marc Schoenauer : -0.05
24 Pierre Dragicevic : -0.05
25 Guillaume Charpiat : -0.05
26 Michel Beaudouin-Lafon : -0.05
27 Fatiha Saïs : -0.04
28 Claude Marché : -0.04
29 Nicolas Bredeche : -0.04
30 Céline Gicquel : -0.03
31 Steven Martin : -0.03
32 Cyril Furtlehner : -0.03
33 Olivier Chapuis : -0.03
34 Philippe Caillou : -0.02
35 Michèle Sebag : -0.02
36 Anastasia Bezerianos : -0.02
37 Isabelle Guyon : -0.02
38 Franck Cappello : -0.01
39 Albert Cohen : -0.01
40 Wendy Mackay : -0.00
41 Raymond Ros : -0.00
42 Olivier Teytaud : -0.00
43 Sébastien Tixeuil : 0.01
44 Nathann Cohen : 0.02
45 Wendy E. Mackay : 0.03
46 Paola Tubaro : 0.09
47 Sarah Cohen-Boulakia : 0.10
dtype: object
VALUE : -0.0506
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.634043
Michèle Sebag 0.415328
Johanne Cohen 0.244186
Albert Cohen 0.560345
Wendy E. Mackay 0.256522
Philippe Caillou 0.597674
Alain Denise 0.297222
Jean-Daniel Fekete 0.514465
Emmanuel Pietriga 0.265079
Yann Ponty 0.590909
Marc Schoenauer 0.505755
Franck Cappello 0.636585
Caroline Appert 0.265217
Michel Beaudouin-Lafon 0.367901
Wendy Mackay 0.340426
Anne Auger 0.535443
Evelyne Lutton 0.356757
Pierre Dragicevic 0.340741
Ioana Manolescu 0.587805
Nikolaus Hansen 0.646914
Nicolas Bredeche 0.597059
Olivier Teytaud 0.572642
François Goasdoué 0.516981
Nathalie Pernelle 0.544118
Fatiha Saïs 0.580488
Sarah Cohen-Boulakia 0.666667
Claude Marché 0.489362
Chantal Reynaud 0.450000
Olivier Chapuis 0.367308
Steven Martin 0.384615
Fatiha Zaidi 0.487500
Balázs Kégl 0.315789
Paola Tubaro 0.758974
Raymond Ros 0.532353
Cyril Furtlehner 0.612821
Anastasia Bezerianos 0.319403
Sylvie Boldo 0.494286
Guillaume Melquiond 0.436364
Marc Baboulin 0.586667
Dimo Brockhoff 0.635897
Nathann Cohen 0.541463
Petra Isenberg 0.453271
Tobias Isenberg 0.576923
Loïc Paulevé 0.721429
Céline Gicquel 0.823684
Isabelle Guyon 0.640449
Guillaume Charpiat 0.506452
Lonni Besançon 0.439394
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.634043 0.010638
Michèle Sebag 0.415328 -0.017518
Johanne Cohen 0.244186 -0.081395
Albert Cohen 0.560345 -0.005172
Wendy E. Mackay 0.256522 0.026087
Philippe Caillou 0.597674 -0.020930
Alain Denise 0.297222 -0.061111
Jean-Daniel Fekete 0.514465 -0.086792
Emmanuel Pietriga 0.265079 -0.098413
Yann Ponty 0.590909 -0.072727
Marc Schoenauer 0.505755 -0.053237
Franck Cappello 0.636585 -0.014634
Caroline Appert 0.265217 -0.076087
Michel Beaudouin-Lafon 0.367901 -0.045679
Wendy Mackay 0.340426 -0.004255
Anne Auger 0.535443 -0.141772
Evelyne Lutton 0.356757 -0.183784
Pierre Dragicevic 0.340741 -0.051852
Ioana Manolescu 0.587805 -0.071951
Nikolaus Hansen 0.646914 -0.114815
Nicolas Bredeche 0.597059 -0.038235
Olivier Teytaud 0.572642 -0.002830
François Goasdoué 0.516981 -0.103774
Nathalie Pernelle 0.544118 -0.073529
Fatiha Saïs 0.580488 -0.041463
Sarah Cohen-Boulakia 0.666667 0.103030
Claude Marché 0.489362 -0.038298
Chantal Reynaud 0.450000 -0.060000
Olivier Chapuis 0.367308 -0.030769
Steven Martin 0.384615 -0.033333
Fatiha Zaidi 0.487500 -0.062500
Balázs Kégl 0.315789 -0.086842
Paola Tubaro 0.758974 0.087179
Raymond Ros 0.532353 -0.002941
Cyril Furtlehner 0.612821 -0.030769
Anastasia Bezerianos 0.319403 -0.016418
Sylvie Boldo 0.494286 -0.062857
Guillaume Melquiond 0.436364 -0.127273
Marc Baboulin 0.586667 -0.064444
Dimo Brockhoff 0.635897 -0.115385
Nathann Cohen 0.541463 0.021951
Petra Isenberg 0.453271 -0.139252
Tobias Isenberg 0.576923 -0.094872
Loïc Paulevé 0.721429 -0.064286
Céline Gicquel 0.823684 -0.034211
Isabelle Guyon 0.640449 -0.015730
Guillaume Charpiat 0.506452 -0.051613
Lonni Besançon 0.439394 -0.084848
Nothing in cache, initial Fitting with min_cluster_size=15 Found 42 clusters in 0.2711373510001067s
Max Fitting with min_cluster_size=30 Found 3 clusters in 0.10500004599998647s
Midpoint Fitting with min_cluster_size=22 Found 4 clusters in 0.10679338699992513s
Re-Fitting with min_cluster_size=15 Found 42 clusters in 0.10345919800010961s
Clusters cached: [3, 4, 42]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 42 clusters in 0.10357241900010195s
Max Fitting with min_cluster_size=30 Found 3 clusters in 0.10556750900013867s
Midpoint Fitting with min_cluster_size=22 Found 4 clusters in 0.10724184199989395s
Re-Fitting with min_cluster_size=15 Found 42 clusters in 0.10214239300012196s
Clusters cached: [3, 4, 42]
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : 0.1325
4_clus__avg_word_couv_0 : 0.6856
4_clus__med_word_couv_0 : 0.6707
4_clus__avg_word_couv_minus_0 : 0.6721
4_clus__big_small_ratio_0 : 72.2667
4_clus__stab_clus_0 : 0.0238
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : 0.1325
4_clus__avg_word_couv_1 : 0.5883
4_clus__med_word_couv_1 : 0.6047
4_clus__avg_word_couv_minus_1 : 0.5757
4_clus__big_small_ratio_1 : 72.2667
4_clus__stab_clus_1 : 0.0000
4_clus__avg_stab_avg : 0.0119
4_clus__avg_couv_avg : 0.6370
4_clus__clu_score : 0.3244
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 visualizations, interaction techniques : s 645...
1 query, ontology : s 252 stb 0.30 + 0.62 - 0.04
2 logic science, verification : s 230 stb 0.40 +...
3 architectures, parallelism : s 222 stb 0.20 + ...
4 networking internet architecture, simulation r...
5 natural language, computation language : s 145...
6 fluid, numerical simulations : s 108 stb 0.00 ...
7 vertices, planar : s 107 stb 0.00 + 0.58 - 0.02
8 secondary structure, bioinformatics : s 106 st...
9 biological, metabolic : s 92 stb 0.00 + 0.66 -...
10 cluster computing, fault : s 82 stb 0.00 + 0.8...
11 neural networks, automl : s 82 stb 0.00 + 0.40...
12 stabilizing, population protocols : s 69 stb 0...
13 humanities social sciences, agent : s 64 stb 0...
14 function evaluations, black optimization : s 5...
15 floating point, arithmetic : s 55 stb 0.00 + 0...
16 optimization control, sphere function : s 54 s...
17 combinatorics, permutations : s 53 stb 0.00 + ...
18 mixed integer linear, numerical results : s 49...
19 monte carlo search, games : s 38 stb 0.00 + 0....
20 french language, signing : s 35 stb 0.00 + 0.6...
21 astrophysics, machine learning challenge : s 3...
22 evolutionary robotics, autonomous : s 32 stb 0...
23 covariance matrix adaptation : s 32 stb 0.00 +...
24 random generation, cellular automata : s 30 st...
25 authors, reviewing : s 30 stb 0.00 + 0.37 - 0.02
26 belief propagation algorithm, condensed matter...
27 discrete event systems, diagnosis : s 28 stb 0...
28 hypermedia, learners : s 26 stb 0.00 + 0.46 - ...
29 multi armed : s 23 stb 0.00 + 0.78 - 0.00
30 quantum : s 22 stb 0.00 + 1.00 - 0.00
31 scientific workflows, scientific workflow : s ...
32 mobile robots : s 20 stb 0.00 + 0.80 - 0.00
33 protein : s 20 stb 0.00 + 1.00 - 0.01
34 treatment, drugs : s 20 stb 0.00 + 0.65 - 0.01
35 twitter, information networks : s 17 stb 0.00 ...
36 electric power, power grids : s 16 stb 0.00 + ...
37 approximate bayesian, genetics : s 16 stb 0.00...
38 omega, occupy : s 15 stb 0.00 + 0.53 - 0.00
39 inconsistency : s 15 stb 0.00 + 0.80 - 0.00
40 image classification, convolutional : s 15 stb...
41 evolutionary computation, genetic algorithms :...
dtype: object
VALUE : 0.6856
4_clus__clus_eval_pos_1_det
0 information visualization, display : s 645 stb...
1 queries, ontologies : s 252 stb 0.00 + 0.49 - ...
2 specification, deductive : s 230 stb 0.00 + 0....
3 compiler, linear systems : s 222 stb 0.00 + 0....
4 networking internet, cloud radio access networ...
5 natural language processing, machine translati...
6 mechanics, flows : s 108 stb 0.00 + 0.57 - 0.02
7 vertex, edges : s 107 stb 0.00 + 0.53 - 0.02
8 secondary, sequence : s 106 stb 0.00 + 0.66 - ...
9 biology, metabolism : s 92 stb 0.00 + 0.59 - 0.02
10 image processing, hyper parameter : s 82 stb 0...
11 fault tolerance, cloud computing : s 82 stb 0....
12 protocols, stabilizing algorithm : s 69 stb 0....
13 social sciences, multiagent systems : s 64 stb...
14 benchmarking, dimension search space : s 57 st...
15 floating point arithmetic, floating : s 55 stb...
16 evolution strategies, noisy optimization : s 5...
17 combinatorial, permutation : s 53 stb 0.00 + 0...
18 integer linear, chance constrained : s 49 stb ...
19 monte carlo, partially observable : s 38 stb 0...
20 motion capture, signs : s 35 stb 0.00 + 0.51 -...
21 cosmic, learning challenge : s 33 stb 0.00 + 0...
22 covariance matrix adaptation evolution, covari...
23 robotics, robotic : s 32 stb 0.00 + 0.94 - 0.01
24 members, reviewers : s 30 stb 0.00 + 0.30 - 0.01
25 words, cellular : s 30 stb 0.00 + 0.63 - 0.03
26 belief propagation, boltzmann machine : s 30 s...
27 distributed discrete event systems, faults : s...
28 forgetting, experimenting : s 26 stb 0.00 + 0....
29 multi armed bandit, multi armed bandits : s 23...
30 quantum physics, circuit : s 22 stb 0.00 + 0.7...
31 workflows : s 22 stb 0.00 + 0.82 - 0.01
32 protein protein, protein complexes : s 20 stb ...
33 patients, clinical : s 20 stb 0.00 + 0.65 - 0.01
34 robots : s 20 stb 0.00 + 0.95 - 0.01
35 tweets, social media : s 17 stb 0.00 + 0.47 - ...
36 network architecture, electricity : s 16 stb 0...
37 demographic, populations : s 16 stb 0.00 + 0.8...
38 inconsistent, description logic : s 15 stb 0.0...
39 experimentally, young : s 15 stb 0.00 + 0.07 -...
40 evolutionary algorithm, coevolution : s 15 stb...
41 monotone, experimentation : s 15 stb 0.00 + 0....
dtype: object
VALUE : 0.5883
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.4585
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.50
2 4_clus__clu_score : 0.32
dtype: object
VALUE : 0.4585
--------- Raw Scores ---------
================================
Run params : {'2_nD__rkey': 'bert', '2_nD__rnum_dims': 768, '2_nD__rnormalize': True, '2_nD__rfamily': 'all-MiniLM-L6-v2', '2_nD__rmax_length': 256}
--------------------------------
Scoring params : {'2_nD__smin_score': 30, '2_nD__srecompute': True, '2_nD__ssample_size': None, '2_nD__sn_neighbors': 10, '2_nD__srandom_state': 42}
================================
----------- Scores -----------
2_nD__neighbors_articles_authors : 0.5509
--------- Desc Scores ---------
2_nD__neighbors_articles_authors_det
0 Céline Gicquel : 0.86
1 Loïc Paulevé : 0.79
2 Nikolaus Hansen : 0.76
3 Dimo Brockhoff : 0.75
4 Anne Auger : 0.68
5 Tobias Isenberg : 0.67
6 Paola Tubaro : 0.67
7 Yann Ponty : 0.66
8 Ioana Manolescu : 0.66
9 Isabelle Guyon : 0.66
10 Franck Cappello : 0.65
11 Marc Baboulin : 0.65
12 Cyril Furtlehner : 0.64
13 Nicolas Bredeche : 0.64
14 Sébastien Tixeuil : 0.62
15 Fatiha Saïs : 0.62
16 François Goasdoué : 0.62
17 Philippe Caillou : 0.62
18 Nathalie Pernelle : 0.62
19 Jean-Daniel Fekete : 0.60
20 Petra Isenberg : 0.59
21 Olivier Teytaud : 0.58
22 Albert Cohen : 0.57
23 Sarah Cohen-Boulakia : 0.56
24 Guillaume Melquiond : 0.56
25 Marc Schoenauer : 0.56
26 Guillaume Charpiat : 0.56
27 Sylvie Boldo : 0.56
28 Fatiha Zaidi : 0.55
29 Evelyne Lutton : 0.54
30 Raymond Ros : 0.54
31 Claude Marché : 0.53
32 Lonni Besançon : 0.52
33 Nathann Cohen : 0.52
34 Chantal Reynaud : 0.51
35 Michèle Sebag : 0.43
36 Steven Martin : 0.42
37 Michel Beaudouin-Lafon : 0.41
38 Balázs Kégl : 0.40
39 Olivier Chapuis : 0.40
40 Pierre Dragicevic : 0.39
41 Emmanuel Pietriga : 0.36
42 Alain Denise : 0.36
43 Wendy Mackay : 0.34
44 Caroline Appert : 0.34
45 Anastasia Bezerianos : 0.34
46 Johanne Cohen : 0.33
47 Wendy E. Mackay : 0.23
Name: 0, dtype: object
VALUE : 0.5509
--------- Raw Scores ---------
2_nD__neighbors_articles_authors
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
Name: 0, dtype: float64
================================
Run params : {'3_2D__rkey': 'umap', '3_2D__rmetric': 'euclidean', '3_2D__rn_neighbors': 20, '3_2D__rmin_dist': 0.5, '3_2D__rinit': 'random', '3_2D__rlearning_rate': 1.0, '3_2D__rn_epochs': None, '3_2D__rrandom_state': None}
--------------------------------
Scoring params : {'3_2D__smin_score': 30, '3_2D__srecompute': True, '3_2D__ssample_size': None, '3_2D__sn_neighbors': 10, '3_2D__srandom_state': 42, '3_2D__smetric': 'euclidean'}
================================
----------- Scores -----------
3_2D__neighbors_articles_authors : 0.4478
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_mean : -0.1031
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_median : -0.0998
3_2D__trustworthiness_sklearn : 0.9004
--------- Desc Scores ---------
3_2D__neighbors_articles_authors_det
0 Loïc Paulevé : 0.80
1 Isabelle Guyon : 0.69
2 Nicolas Bredeche : 0.64
3 Paola Tubaro : 0.63
4 Céline Gicquel : 0.63
5 Nikolaus Hansen : 0.62
6 Dimo Brockhoff : 0.62
7 Yann Ponty : 0.58
8 Sébastien Tixeuil : 0.57
9 Ioana Manolescu : 0.56
10 Olivier Teytaud : 0.56
11 Sarah Cohen-Boulakia : 0.55
12 Sylvie Boldo : 0.53
13 Raymond Ros : 0.53
14 François Goasdoué : 0.53
15 Marc Baboulin : 0.52
16 Cyril Furtlehner : 0.52
17 Anne Auger : 0.51
18 Fatiha Saïs : 0.50
19 Marc Schoenauer : 0.49
20 Nathann Cohen : 0.48
21 Albert Cohen : 0.47
22 Jean-Daniel Fekete : 0.46
23 Claude Marché : 0.46
24 Guillaume Charpiat : 0.43
25 Tobias Isenberg : 0.43
26 Nathalie Pernelle : 0.42
27 Franck Cappello : 0.42
28 Petra Isenberg : 0.41
29 Chantal Reynaud : 0.40
30 Guillaume Melquiond : 0.40
31 Lonni Besançon : 0.39
32 Fatiha Zaidi : 0.37
33 Philippe Caillou : 0.37
34 Michèle Sebag : 0.35
35 Steven Martin : 0.34
36 Olivier Chapuis : 0.32
37 Michel Beaudouin-Lafon : 0.31
38 Wendy Mackay : 0.30
39 Anastasia Bezerianos : 0.29
40 Caroline Appert : 0.29
41 Pierre Dragicevic : 0.28
42 Balázs Kégl : 0.27
43 Evelyne Lutton : 0.27
44 Alain Denise : 0.26
45 Wendy E. Mackay : 0.26
46 Emmanuel Pietriga : 0.25
47 Johanne Cohen : 0.20
dtype: object
VALUE : 0.4478
3_2D__2_nD_3_2D_neighbors_articles_authors_dim_det
0 Evelyne Lutton : -0.27
1 Philippe Caillou : -0.25
2 Tobias Isenberg : -0.25
3 Franck Cappello : -0.23
4 Céline Gicquel : -0.23
5 Nathalie Pernelle : -0.19
6 Fatiha Zaidi : -0.18
7 Petra Isenberg : -0.18
8 Anne Auger : -0.17
9 Guillaume Melquiond : -0.16
10 Jean-Daniel Fekete : -0.14
11 Nikolaus Hansen : -0.14
12 Dimo Brockhoff : -0.14
13 Lonni Besançon : -0.13
14 Johanne Cohen : -0.13
15 Guillaume Charpiat : -0.13
16 Balázs Kégl : -0.13
17 Marc Baboulin : -0.13
18 Cyril Furtlehner : -0.13
19 Fatiha Saïs : -0.12
20 Emmanuel Pietriga : -0.12
21 Chantal Reynaud : -0.11
22 Pierre Dragicevic : -0.11
23 Michel Beaudouin-Lafon : -0.10
24 Alain Denise : -0.10
25 Albert Cohen : -0.10
26 Ioana Manolescu : -0.10
27 François Goasdoué : -0.09
28 Yann Ponty : -0.08
29 Olivier Chapuis : -0.08
30 Steven Martin : -0.08
31 Michèle Sebag : -0.08
32 Claude Marché : -0.07
33 Marc Schoenauer : -0.07
34 Caroline Appert : -0.05
35 Sébastien Tixeuil : -0.05
36 Wendy Mackay : -0.05
37 Anastasia Bezerianos : -0.04
38 Nathann Cohen : -0.04
39 Paola Tubaro : -0.04
40 Sylvie Boldo : -0.02
41 Olivier Teytaud : -0.02
42 Sarah Cohen-Boulakia : -0.01
43 Raymond Ros : -0.00
44 Nicolas Bredeche : 0.00
45 Loïc Paulevé : 0.02
46 Wendy E. Mackay : 0.03
47 Isabelle Guyon : 0.04
dtype: object
VALUE : -0.1031
--------- Raw Scores ---------
3_2D__neighbors_articles_authors
Sébastien Tixeuil 0.570213
Michèle Sebag 0.354745
Johanne Cohen 0.195349
Albert Cohen 0.468966
Wendy E. Mackay 0.258696
Philippe Caillou 0.369767
Alain Denise 0.261111
Jean-Daniel Fekete 0.461635
Emmanuel Pietriga 0.247619
Yann Ponty 0.580303
Marc Schoenauer 0.491367
Franck Cappello 0.421951
Caroline Appert 0.286957
Michel Beaudouin-Lafon 0.311111
Wendy Mackay 0.295745
Anne Auger 0.505063
Evelyne Lutton 0.267568
Pierre Dragicevic 0.282716
Ioana Manolescu 0.563415
Nikolaus Hansen 0.623457
Nicolas Bredeche 0.635294
Olivier Teytaud 0.558491
François Goasdoué 0.528302
Nathalie Pernelle 0.423529
Fatiha Saïs 0.504878
Sarah Cohen-Boulakia 0.551515
Claude Marché 0.457447
Chantal Reynaud 0.400000
Olivier Chapuis 0.315385
Steven Martin 0.338462
Fatiha Zaidi 0.371875
Balázs Kégl 0.273684
Paola Tubaro 0.633333
Raymond Ros 0.532353
Cyril Furtlehner 0.517949
Anastasia Bezerianos 0.291045
Sylvie Boldo 0.534286
Guillaume Melquiond 0.400000
Marc Baboulin 0.522222
Dimo Brockhoff 0.615385
Nathann Cohen 0.480488
Petra Isenberg 0.414953
Tobias Isenberg 0.425641
Loïc Paulevé 0.802381
Céline Gicquel 0.631579
Isabelle Guyon 0.692135
Guillaume Charpiat 0.429032
Lonni Besançon 0.393939
dtype: float64
3_2D__2_nD_3_2D_neighbors_articles_authors_dim
2_nD__neighbors_articles_authors \
Sébastien Tixeuil 0.623404
Michèle Sebag 0.432847
Johanne Cohen 0.325581
Albert Cohen 0.565517
Wendy E. Mackay 0.230435
Philippe Caillou 0.618605
Alain Denise 0.358333
Jean-Daniel Fekete 0.601258
Emmanuel Pietriga 0.363492
Yann Ponty 0.663636
Marc Schoenauer 0.558993
Franck Cappello 0.651220
Caroline Appert 0.341304
Michel Beaudouin-Lafon 0.413580
Wendy Mackay 0.344681
Anne Auger 0.677215
Evelyne Lutton 0.540541
Pierre Dragicevic 0.392593
Ioana Manolescu 0.659756
Nikolaus Hansen 0.761728
Nicolas Bredeche 0.635294
Olivier Teytaud 0.575472
François Goasdoué 0.620755
Nathalie Pernelle 0.617647
Fatiha Saïs 0.621951
Sarah Cohen-Boulakia 0.563636
Claude Marché 0.527660
Chantal Reynaud 0.510000
Olivier Chapuis 0.398077
Steven Martin 0.417949
Fatiha Zaidi 0.550000
Balázs Kégl 0.402632
Paola Tubaro 0.671795
Raymond Ros 0.535294
Cyril Furtlehner 0.643590
Anastasia Bezerianos 0.335821
Sylvie Boldo 0.557143
Guillaume Melquiond 0.563636
Marc Baboulin 0.651111
Dimo Brockhoff 0.751282
Nathann Cohen 0.519512
Petra Isenberg 0.592523
Tobias Isenberg 0.671795
Loïc Paulevé 0.785714
Céline Gicquel 0.857895
Isabelle Guyon 0.656180
Guillaume Charpiat 0.558065
Lonni Besançon 0.524242
3_2D__neighbors_articles_authors scdec
Sébastien Tixeuil 0.570213 -0.053191
Michèle Sebag 0.354745 -0.078102
Johanne Cohen 0.195349 -0.130233
Albert Cohen 0.468966 -0.096552
Wendy E. Mackay 0.258696 0.028261
Philippe Caillou 0.369767 -0.248837
Alain Denise 0.261111 -0.097222
Jean-Daniel Fekete 0.461635 -0.139623
Emmanuel Pietriga 0.247619 -0.115873
Yann Ponty 0.580303 -0.083333
Marc Schoenauer 0.491367 -0.067626
Franck Cappello 0.421951 -0.229268
Caroline Appert 0.286957 -0.054348
Michel Beaudouin-Lafon 0.311111 -0.102469
Wendy Mackay 0.295745 -0.048936
Anne Auger 0.505063 -0.172152
Evelyne Lutton 0.267568 -0.272973
Pierre Dragicevic 0.282716 -0.109877
Ioana Manolescu 0.563415 -0.096341
Nikolaus Hansen 0.623457 -0.138272
Nicolas Bredeche 0.635294 0.000000
Olivier Teytaud 0.558491 -0.016981
François Goasdoué 0.528302 -0.092453
Nathalie Pernelle 0.423529 -0.194118
Fatiha Saïs 0.504878 -0.117073
Sarah Cohen-Boulakia 0.551515 -0.012121
Claude Marché 0.457447 -0.070213
Chantal Reynaud 0.400000 -0.110000
Olivier Chapuis 0.315385 -0.082692
Steven Martin 0.338462 -0.079487
Fatiha Zaidi 0.371875 -0.178125
Balázs Kégl 0.273684 -0.128947
Paola Tubaro 0.633333 -0.038462
Raymond Ros 0.532353 -0.002941
Cyril Furtlehner 0.517949 -0.125641
Anastasia Bezerianos 0.291045 -0.044776
Sylvie Boldo 0.534286 -0.022857
Guillaume Melquiond 0.400000 -0.163636
Marc Baboulin 0.522222 -0.128889
Dimo Brockhoff 0.615385 -0.135897
Nathann Cohen 0.480488 -0.039024
Petra Isenberg 0.414953 -0.177570
Tobias Isenberg 0.425641 -0.246154
Loïc Paulevé 0.802381 0.016667
Céline Gicquel 0.631579 -0.226316
Isabelle Guyon 0.692135 0.035955
Guillaume Charpiat 0.429032 -0.129032
Lonni Besançon 0.393939 -0.130303
Nothing in cache, initial Fitting with min_cluster_size=15 Found 4 clusters in 0.29130253199991785s
No need Re-Fitting with min_cluster_size=15
Clusters cached: [4]
Nothing in cache, initial Fitting with min_cluster_size=15 Found 4 clusters in 0.10657290900007865s
No need Re-Fitting with min_cluster_size=15
Clusters cached: [4]
================================
Run params : {'4_clus__rkey': 'hdbscan', '4_clus__rbase_factor': 3}
--------------------------------
Scoring params : {'4_clus__siter_stab': 2, '4_clus__sremove_stab': [0, 0.01, 0.03, 0.1, 0.25], '4_clus__smetric': 'euclidean', '4_clus__srandom_state': None}
================================
----------- Scores -----------
4_clus__nb_clust_0 : 8.0000
4_clus__silhouette_0 : -0.0674
4_clus__avg_word_couv_0 : 0.5945
4_clus__med_word_couv_0 : 0.7148
4_clus__avg_word_couv_minus_0 : 0.5878
4_clus__big_small_ratio_0 : 204.0500
4_clus__stab_clus_0 : 0.0000
4_clus__nb_clust_1 : 24.0000
4_clus__silhouette_1 : -0.0674
4_clus__avg_word_couv_1 : 0.5816
4_clus__med_word_couv_1 : 0.6955
4_clus__avg_word_couv_minus_1 : 0.5751
4_clus__big_small_ratio_1 : 204.0500
4_clus__stab_clus_1 : 0.0000
4_clus__avg_stab_avg : 0.0000
4_clus__avg_couv_avg : 0.5881
4_clus__clu_score : 0.2940
--------- Desc Scores ---------
4_clus__clus_eval_pos_0_det
0 logic science, cluster computing : s 4081 stb ...
1 fluid, numerical simulations : s 108 stb 0.00 ...
2 energy physics, machine learning challenge : s...
3 astrophysics : s 20 stb 0.00 + 0.80 - 0.00
dtype: object
VALUE : 0.5945
4_clus__clus_eval_pos_1_det
0 discrete, queries : s 4081 stb 0.00 + 0.09 - 0.00
1 mechanics, flows : s 108 stb 0.00 + 0.56 - 0.02
2 large hadron collider, learning challenge : s ...
3 cosmic, sciences universe : s 20 stb 0.00 + 0....
dtype: object
VALUE : 0.5816
--------- Raw Scores ---------
['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']
================================
Run params : {}
--------------------------------
Scoring params : {'6_pst__sname_list': ['2_nD__neighbors_articles_authors', '3_2D__neighbors_articles_authors', '4_clus__clu_score']}
================================
----------- Scores -----------
6_pst__final_score : 0.4309
--------- Desc Scores ---------
6_pst__final_score_det
0 2_nD__neighbors_articles_authors : 0.55
1 3_2D__neighbors_articles_authors : 0.45
2 4_clus__clu_score : 0.29
dtype: object
VALUE : 0.4309
--------- Raw Scores ---------
<cartodata.model_selection.utils.Results object at 0x7efd56e60c70>
.. GENERATED FROM PYTHON SOURCE LINES 194-195
We can see that we have run the first 6 parameter sets in the dataframe.
.. GENERATED FROM PYTHON SOURCE LINES 195-197
.. code-block:: Python
len(experiment.results.runs_)
.. rst-class:: sphx-glr-script-out
.. code-block:: none
6
.. rst-class:: sphx-glr-timing
**Total running time of the script:** (28 minutes 16.209 seconds)
.. _sphx_glr_download_auto_examples_experiment_pipeline_lisn.py:
.. only:: html
.. container:: sphx-glr-footer sphx-glr-footer-example
.. container:: sphx-glr-download sphx-glr-download-jupyter
:download:`Download Jupyter notebook: experiment_pipeline_lisn.ipynb <experiment_pipeline_lisn.ipynb>`
.. container:: sphx-glr-download sphx-glr-download-python
:download:`Download Python source code: experiment_pipeline_lisn.py <experiment_pipeline_lisn.py>`
.. container:: sphx-glr-download sphx-glr-download-zip
:download:`Download zipped: experiment_pipeline_lisn.zip <experiment_pipeline_lisn.zip>`
.. only:: html
.. rst-class:: sphx-glr-signature
`Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_