Skip to content

Broundal/Pytolemaic

Repository files navigation

PyPI - Version Unittests PyPI - License

Pytolemaic

What is Pytolemaic

Pytolemaic package analyzes your model and dataset and measure their quality.

The package supports classification/regression models built for tabular datasets (e.g. sklearn's regressors/classifiers), but will also support custom made models as long as they implement sklearn's API.

The package is aimed for personal use and comes with no guarantees. I hope you will find it useful. I will appreciate any feedback you have.

Install

pip install pytolemaic

Basic usage

from pytolemaic import PyTrust

pytrust = PyTrust(model=estimator,
                  xtrain=xtrain, ytrain=ytrain,
                  xtest=xtest, ytest=ytest)
   
# run all analysis and print insights:,
insights = pytrust.insights()
print("\n".join(insights))

# run analysis and plot graphs
pytrust.plot()

supported features

The package contains the following functionalities:

On model creation

  • Dataset Analysis: Analysis aimed to detect issues in the dataset.
  • Sensitivity Analysis: Calculation of feature importance for given model, either via sensitivity to feature value or sensitivity to missing values.
  • Vulnerability report: Based on the feature sensitivity we measure model's vulnerability in respect to imputation, leakage, and # of features.
  • Scoring report: Report model's score on test data with confidence interval.
  • separation quality: Measure whether train and test data comes from the same distribution.
  • Overall quality: Provides overall quality measures

On prediction

  • Prediction uncertainty: Provides an uncertainty measure for given model's prediction.
  • Lime explanation: Provides Lime explanation for sample of interest.

How to use:

Get started by calling help() function (Recommended!):

   from pytolemaic import help
   supported_keys = help()
   # or
   help(key='basic usage')

Example for performing all available analysis with PyTrust:

   from pytolemaic import PyTrust

   pytrust = PyTrust(
       model=estimator,
       xtrain=xtrain, ytrain=ytrain,
       xtest=xtest, ytest=ytest)
       
   # run all analysis and get a list of distilled insights",
   insights = pytrust.insights()
   print("\n".join(insights))
    
   # run all analysis and plot all graphs
   pytrust.plot()
   
   # print all data gathered
   import pprint
   pprint(report.to_dict(printable=True))

In case of need to access only specific analysis (usually to save time)

   # dataset analysis report
   dataset_analysis_report = pytrust.dataset_analysis_report
   
   # feature sensitivity report
   sensitivity_report = pytrust.sensitivity_report
   
   # model's performance report
   scoring_report = pytrust.scoring_report
   
   # overall model's quality report
   quality_report = pytrust.quality_report
   
   # with any of the above reports
   report = <desired report>
   print("\n".join(report.insights()))
   
   report.plot() # plot graphs
   pprint(report.to_dict(printable=True)) # export report as a dictionary
   pprint(report.to_dict_meaning()) # print documentation for above dictionary
          

Analysis of predictions

 
   # estimate uncertainty of a prediction
   uncertainty_model = pytrust.create_uncertainty_model()
   
   # explain a prediction with Lime
   create_lime_explainer = pytrust.create_lime_explainer()
   

Examples on toy dataset can be found in /examples/toy_examples/ Examples on 'real-life' datasets can be found in /examples/interesting_examples/

Output examples:

Sensitivity Analysis:

  • The sensitivity of each feature ([0,1], normalized to sum of 1):
 'sensitivity_report': {
    'method': 'shuffled',
    'sensitivities': {
        'age': 0.12395,
        'capital-gain': 0.06725,
        'capital-loss': 0.02465,
        'education': 0.05769,
        'education-num': 0.13765,
        ...
      }
  }
  • Simple statistics on the feature sensitivity:
'shuffle_stats_report': {
     'n_features': 14,
     'n_low': 1,
     'n_zero': 0
}
  • Naive vulnerability scores ([0,1], lower is better):

    • Imputation: sensitivity of the model to missing values.
    • Leakge: chance of the model to have leaking features.
    • Too many features: Whether the model is based on too many features.
'vulnerability_report': {
     'imputation': 0.35,
     'leakage': 0,
     'too_many_features': 0.14
}  

scoring report

For given metric, the score and confidence intervals (CI) is calculated

'recall': {
    'ci_high': 0.763,
    'ci_low': 0.758,
    'ci_ratio': 0.023,
    'metric': 'recall',
    'value': 0.760,
},
'auc': {
    'ci_high': 0.909,
    'ci_low': 0.907,
    'ci_ratio': 0.022,
    'metric': 'auc',
    'value': 0.907
}    

Additionally, score quality measures the quality of the score based on the separability (auc score) between train and test sets.

Value of 1 means test set has same distribution as train set. Value of 0 means test set has fundamentally different distribution.

'separation_quality': 0.00611         

Combining the above measures into a single number we provide the overall quality of the model/dataset.

Higher quality value ([0,1]) means better dataset/model.

quality_report : { 
'model_quality_report': {
   'model_loss': 0.24,
   'model_quality': 0.41,
   'vulnerability_report': {...}},
   
'test_quality_report': {
   'ci_ratio': 0.023, 
   'separation_quality': 0.006, 
   'test_set_quality': 0},
   
'train_quality_report': {
   'train_set_quality': 0.85,
   'vulnerability_report': {...}}
  

prediction uncertainty

The module can be used to yield uncertainty measure for predictions.

    uncertainty_model = pytrust.create_uncertainty_model(method='confidence')
    predictions = uncertainty_model.predict(x_pred) # same as model.predict(x_pred)
    uncertainty = uncertainty_model.uncertainty(x_pred)

Lime explanation

The module can be used to produce lime explanations for sample of interest.

    explainer = pytrust.create_lime_explainer()
    explainer.explain(sample) # returns a dictionary
    explainer.plot(sample) # produce a graphical explanation