Note
Go to the end to download the full example code
Columns Info#
This notebook provides an overview for using and understanding the columns info check.
Structure:
What are columns info#
The ColumnsInfo check returns the role and logical type of each column (e.g. date, categorical, numerical etc.).
Imports#
import numpy as np
import pandas as pd
from deepchecks.tabular import Dataset
from deepchecks.tabular.checks import ColumnsInfo
Generating data#
num_fe = np.random.rand(500)
cat_fe = np.random.randint(3, size=500)
num_col = np.random.rand(500)
date = range(1635693229, 1635693729)
index = range(500)
data = {'index': index, 'date': date, 'a': cat_fe, 'b': num_fe, 'c': num_col, 'label': cat_fe}
df = pd.DataFrame.from_dict(data)
dataset = Dataset(df, label='label', datetime_name='date', index_name='index', features=['a', 'b'], cat_features=['a'])
Running columns info check#
check = ColumnsInfo()
Total running time of the script: (0 minutes 0.022 seconds)