Residuals#

Objectives: what you will take away#

  • How-To Retrieve global and local residuals.

Prerequisites: before you begin#

Data#

Our example dataset for this recipe is the well known Adult dataset. It is accessible via the pmlb package installed earlier. We use the fetch_data() function to retrieve the dataset in Step 1 below.

Concepts & Terminology#

How-To Guide#

Setup#

The user guide assumes you have created and setup a Trainee as demonstrated in basic workflow. The created Trainee will be referenced as trainee in the sections below.

[1]:

import pandas as pd
from pmlb import fetch_data

from howso.engine import Trainee
from howso.utilities import infer_feature_attributes

df = fetch_data('adult').sample(1_000)
features = infer_feature_attributes(df)

trainee = Trainee(features=features)
trainee.train(df)
trainee.analyze()

features.to_dataframe()

/home/docs/checkouts/readthedocs.org/user_builds/diveplane-howso-docs/envs/latest/lib/python3.11/site-packages/howso/utilities/feature_attributes/pandas.py:145: UserWarning: You have one or more suggestions to consider for your feature attributes configuration. Please view them by printing the `suggestions` property of your returned feature attributes object (`your_attributes_object.suggestions`).
  warnings.warn(suggestion_warning, UserWarning)

[1]:
type decimal_places bounds data_type original_type
min max allow_null observed_min observed_max nulls_observed data_type size
age continuous 0 0.0 137.0 True 17.0 90.0 False number numeric 8
workclass nominal 0 NaN NaN True NaN NaN False number integer 8
fnlwgt continuous 0 0.0 1944119.0 True 13862.0 1184622.0 False number numeric 8
education nominal 0 NaN NaN True NaN NaN False number integer 8
education-num continuous 0 0.0 26.0 True 1.0 16.0 False number numeric 8
marital-status nominal 0 NaN NaN True NaN NaN False number integer 8
occupation nominal 0 NaN NaN True NaN NaN False number integer 8
relationship nominal 0 NaN NaN True NaN NaN False number integer 8
race nominal 0 NaN NaN True NaN NaN False number integer 8
sex nominal 0 NaN NaN True NaN NaN False number integer 8
capital-gain continuous 0 0.0 164870.0 True 0.0 99999.0 False number numeric 8
capital-loss continuous 0 0.0 3982.0 True 0.0 2415.0 False number numeric 8
hours-per-week continuous 0 0.0 163.0 True 1.0 99.0 False number numeric 8
native-country continuous 0 0.0 66.0 True 0.0 40.0 False number integer 8
target nominal 0 NaN NaN True NaN NaN False number integer 8

Local Residuals#

Local metrics are retrieved through using Trainee.react(). Both Robust and non-robust (full) versions are available, although full is recommended for residuals.

[2]:

# Get local full residuals
details = {'feature_full_residuals_for_case': True}
results = trainee.react(
    df.iloc[[-1]],
    context_features=features.get_names(without=["target"]),
    action_features=["target"],
    details=details
)

residuals = results['details']['feature_full_residuals_for_case']
residuals

[2]:
workclass sex race native-country education-num occupation education marital-status capital-loss fnlwgt target age capital-gain relationship hours-per-week
0 [0.17753964974126968, 0] [0.359369678865173, 0] [0.13143175714420807, 0] [3.2401110524445897, 0] [1.7763568394002505e-15, 0] [0.7372711123793414, 0] [0, 0] [0, 0] [0, 0] [191808.99970349576, 0] [0, 0] [2.649609932850961, 0] [0, 0] [0.3934297424337464, 0] [0.40326033227631797, 0]

Global Residuals#

Howso has the ability to retrieve both local vs global metrics. Global metrics are retrieved through using Trainee.react_aggregate(). Both Robust and non-robust (full) versions are also available.

[3]:

# Get global full residuals
residuals = trainee.react_aggregate(
    details={'feature_full_residuals': True},
).to_dataframe()
residuals

[3]:
workclass sex race native-country education-num occupation education marital-status target fnlwgt capital-loss capital-gain age relationship hours-per-week
feature_full_residuals [0.3589348618269888, 0] [0.2417996938574392, 0] [0.2397361995437431, 0] [2.749922760136048, 0] [0.05464532505609074, 0] [0.837073721474876, 0] [0.005077383747108766, 0] [0.22988024313953254, 0] [0.21531142446852702, 0] [82894.50674566082, 0] [83.20948637979849, 0] [1615.258946011626, 0] [9.380478286325173, 0] [0.32641103508447417, 0] [8.060335593020671, 0]

API References#