def test_transform_hour_fmt_list(self):
df = pd.DataFrame({
"date_column": [
"2016-01-01 15:16:45",
"2016-06-28 12:18:51",
"2016-07-28 01:01:01",
]
})
trainable = Map(columns=[hour(it.date_column, "%Y-%m-%d %H:%M:%S")])
trained = trainable.fit(df)
transformed_df = trained.transform(df)
self.assertEqual(transformed_df["date_column"][0], 15)
self.assertEqual(transformed_df["date_column"][1], 12)
self.assertEqual(transformed_df["date_column"][2], 1)
def test_transform_hour_fmt_list(self):
df = pd.DataFrame({
"date_column": [
"2016-01-01 15:16:45",
"2016-06-28 12:18:51",
"2016-07-28 01:01:01",
]
})
sdf = self.sqlCtx.createDataFrame(df)
trainable = Map(columns=[hour(it.date_column, "y-M-d HH:mm:ss")])
trained = trainable.fit(sdf)
transformed_df = trained.transform(sdf)
self.assertEqual(transformed_df.collect()[0]["date_column"], 15)
self.assertEqual(transformed_df.collect()[1]["date_column"], 12)
self.assertEqual(transformed_df.collect()[2]["date_column"], 1)