First Example with Modeva

Installation

# To install the required package, use the following command:
# !pip install modeva

Authentication

# To get authentication, use the following command: (To get full access please replace the token to your own token)
# from modeva.utils.authenticate import authenticate
# authenticate(auth_code='eaaa4301-b140-484c-8e93-f9f633c8bacb')

Import modeva modules.

from modeva import DataSet
from modeva import TestSuite
from modeva.models import MoXGBRegressor, MoLGBMRegressor

Load BikeSharing Dataset

ds = DataSet()
ds.load(name="BikeSharing")
ds.set_random_split()

✓ Auth code found in local storage.
Authenticating Modeva...
✓ License is active and valid.
✓ Authenticated successfully!

Fit XGB and LGBM models

model1 = MoXGBRegressor(name="XGB")
model1.fit(ds.train_x, ds.train_y)

model2 = MoLGBMRegressor(name="LGBM-2", max_depth=2, verbose=-1)
model2.fit(ds.train_x, ds.train_y)

MoLGBMRegressor(boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,
                importance_type='split', learning_rate=0.1, max_depth=2,
                min_child_samples=20, min_child_weight=0.001,
                min_split_gain=0.0, n_estimators=100, n_jobs=None,
                num_leaves=31, objective=None, random_state=None, reg_alpha=0.0,
                reg_lambda=0.0, subsample=1.0, subsample_for_bin=200000,
                subsample_freq=0, verbose=-1)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Model Explainability (PDP for hr)

ts = TestSuite(ds, model1)
results = ts.explain_pdp("hr")
results.plot()

Model Explainability (PDP for season)

results = ts.explain_pdp("season")
results.plot()

Diagnostics (accuracy)

results = ts.diagnose_accuracy_table()
results.table

	MSE	MAE	R2
train	803.002976	18.482872	0.975505
test	1718.229803	26.110423	0.948508
GAP	915.226826	7.627551	-0.026996

Diagnostics (slicing accuracy)

results = ts.diagnose_slicing_accuracy(features=(("hr", ), ("season", )), method="uniform",
                                       bins=10, metric="MSE")
results.plot()

Model comparison (slicing accuracy)

tsc = TestSuite(ds, models=[model1, model2])
results = tsc.compare_slicing_accuracy(features="hr", method="quantile",
                                       bins=10, metric="MSE")
results.plot()