regression-inference 1.1.1

OLS and MLE regression inference for Python

regression-inference

https://pypi.org/project/regression-inference/

pip install regression-inference

Python packaged designed for inference workflows using MLE and OLS.

Usage

Import all utilities:

from regression_inference import *

Import select utilities:

from regression_inference import LinearRegression, LogisticRegression, summary

Documentation

See the provided notebooks for example workflows.

/tests/notebooks/linear_regression_example.ipynb

/tests/notebooks/logit_regression_example.ipynb

Output Example

Stacked outputs using summary

print(summary(model, robust_model))

==================================================
OLS Regression Results
--------------------------------------------------
Dependent:                     educ    robust educ
--------------------------------------------------
 
const                     7.3256***      7.3256***
                           (0.3684)       (0.4345)
 
paeduc                    0.2144***      0.2144***
                           (0.0241)       (0.0236)
 
maeduc                    0.2569***      0.2569***
                           (0.0271)       (0.0294)
 
age                       0.0241***      0.0241***
                           (0.0043)       (0.0042)

--------------------------------------------------
R-squared                     0.276          0.276
Adjusted R-squared            0.274          0.274
F Statistic                 177.548        177.548
Observations               1402.000       1402.000
Log Likelihood            -3359.107      -3359.107
AIC                        6726.213       6726.213
BIC                        6747.196       6747.196
TSS                       13663.270      13663.270
RSS                        9893.727       9893.727
ESS                        3769.543       3769.543
MSE                           7.077          7.077
==================================================
*p<0.1; **p<0.05; ***p<0.01

Logistic Regression Summary

===================================
Logistic Regression Results
-----------------------------------
Dependent:                    GRADE
-----------------------------------
 
const                    -13.0213**
                           (5.1976)
 
GPA                        2.8261**
                           (1.2675)
 
TUCE                         0.0952
                           (0.1179)
 
PSI                        2.3787**
                           (0.9644)

-----------------------------------
Pseudo R-squared              0.374
LR Statistic                 15.404
Observations                 32.000
Log Likelihood              -12.890
Deviance                     25.779
Null Deviance                41.183
AIC                          33.779
BIC                          39.642
===================================
*p<0.1; **p<0.05; ***p<0.01

Coefficient Inference Table

Generate an inference table on fitted model objects.

The inference table can be converted to a pd.DataFrame object.

pd.DataFrame(model.inference_table())

[Out]: [{'feature': 'const',
         'coefficient': np.float64(7.3256),
         'std_error': np.float64(0.3684),
         't_statistic': np.float64(19.887),
         'P>|t|': '0.000',
         'ci_low_0.05': np.float64(6.603),
         'ci_high_0.05': np.float64(8.048)},
        {'feature': 'paeduc',
         'coefficient': np.float64(0.2144),
         'std_error': np.float64(0.0241),
         't_statistic': np.float64(8.8796),
         'P>|t|': '0.000',
         'ci_low_0.05': np.float64(0.167),
         'ci_high_0.05': np.float64(0.262)},
        {'feature': 'maeduc',
         'coefficient': np.float64(0.2569),
         'std_error': np.float64(0.0271),
         't_statistic': np.float64(9.4725),
         'P>|t|': '0.000',
         'ci_low_0.05': np.float64(0.204),
         'ci_high_0.05': np.float64(0.31)},
        {'feature': 'age',
         'coefficient': np.float64(0.0241),
         'std_error': np.float64(0.0043),
         't_statistic': np.float64(5.5789),
         'P>|t|': '0.000',
         'ci_low_0.05': np.float64(0.016),
         'ci_high_0.05': np.float64(0.033)}]

Predictions

Extract the order of feature names using feature_names[:1]

model.feature_names[1:]

[Out]: Index(['paeduc', 'maeduc', 'age'], dtype='object')

Predict in the order of the feature names.

model.predict(np.array([[0, 0, 0], ]))

[Out]: array([7.32564767])

Advanced Predictions

Use iterations to make predictions over a discrete range of values

Use return_table = True to generate a dictionary of prediction statistics instead of an array of values.

prediction_set = [
    (np.array([[i, X['maeduc'].mean(), X['age'].mean()],]))
    for i in range(int(X['paeduc'].min()), int(X['paeduc'].max())+1)
    ] 
predictions = pd.concat([pd.DataFrame(model.predict(i, return_table=True)) for i in prediction_set], ignore_index=True)
predictions

Predictions on a Logistic Regression Model

prediction_set = [
    np.array([[2.66, 20.0, 0.0]]),
    np.array([[2.89, 22.0, 0.0]]),
    np.array([[3.28, 24.0, 0.0]]),
    np.array([[2.92, 12.0, 0.0]]),
]
predictions = pd.concat([pd.DataFrame(model.predict(test_set, return_table=True)) for test_set in prediction_set], ignore_index=True)
predictions

Variance Inflation Factor

Variance Inflation Factor can be generated for the model's features.

model.variance_inflation_factor()

Dictionary output can be converted into a pd.DataFrame object

{'feature': Index(['paeduc', 'maeduc', 'age'], dtype='object'),
 'VIF': array([2.0233, 2.0285, 1.0971])}