Book a Demo!
CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
ethen8181
GitHub Repository: ethen8181/machine-learning
Path: blob/master/data_science_is_software/tests/test_example.py
2574 views
1
import os
2
import pytest
3
import pandas as pd
4
5
@pytest.fixture()
6
def df():
7
"""read in the raw data file and return the dataframe"""
8
pump_data_path = os.path.join( 'data', 'raw', 'pumps_train_values.csv' )
9
df = pd.read_csv(pump_data_path)
10
return df
11
12
13
def test_df_fixture(df):
14
assert df.shape == (59400, 40)
15
16
useful_columns = [ 'amount_tsh', 'gps_height', 'longitude', 'latitude', 'region',
17
'population', 'construction_year', 'extraction_type_class',
18
'management_group', 'quality_group', 'source_type',
19
'waterpoint_type', 'status_group' ]
20
21
for column in useful_columns:
22
assert column in df.columns
23
24