PNDSPY1007¶

Message pandas.core.series.Series.str.slice has a partial mapping because it has several scenarios not supported in Snowpark pandas.

Category Warning

Description¶

This issue appears when the SMA identifies a pandas.core.series.Series.str.slice usage. Snowpark pandas offers a partial equivalent, but the current implementation has two unsupported scenarios.

Scenarios¶

The following scenarios illustrate the two unsupported use cases.

Scenario 1¶

The first scenario is when the method receives mixed data type columns, which causes it to behave unexpectedly. All values within a column must be of the same type.

Input¶

The following example shows an unsupported use of pandas.core.series.Series.str.slice.

import pandas as pd

s = pd.Series(["String", (1, 2, 3), ["a", "b", "c"], 123, -456, {1: "Hello", "2": "World"}])
print(s.str.slice(1))

Copy

Output¶

The SMA adds the EWI PNDSPY1007 to the output code to indicate that it has a scenario not supported in Snowpark pandas.

from snowflake.snowpark.modin import plugin
import modin.pandas as pd

s = pd.Series(["String", (1, 2, 3), ["a", "b", "c"], 123, -456, {1: "Hello", "2": "World"}])
#EWI: PNDSPY1007 => pandas.core.series.Series.str.slice has a partial mapping,because has several scenarios not supported in Snowpark pandas.
print(s.str.slice(1))

Copy

Recommended fix¶

Ensure that the Series contains only one type of data (all strings, all lists, or all dicts). No code change is strictly required, but be aware that this operation might not work as expected in Snowpark pandas.

Scenario 2¶

The second scenario is when a column contains list values and the step parameter is set to a value other than one.

Input¶

The following example shows an unsupported use of pandas.core.series.Series.str.slice.

import pandas as pd

ser = pd.Series(["koala", "dog", "chameleon","cat", "mouse", "elephant","lion", "tiger", "bear"])
print(ser.str.slice(step=3))

Copy

Output¶

The SMA adds the EWI PNDSPY1007 to the output code to indicate that it has a scenario not supported in Snowpark pandas.

from snowflake.snowpark.modin import plugin
import modin.pandas as pd

ser = pd.Series(["koala", "dog", "chameleon","cat", "mouse", "elephant","lion", "tiger", "bear"])
#EWI: PNDSPY1007 => pandas.core.series.Series.str.slice has a partial mapping, because has several scenarios not supported in Snowpark pandas.
print(ser.str.slice(step=3))

Copy

Recommended fix¶

This requires a manual change. Use the apply function to achieve the same behavior with a lambda function. Here is the previous output code with the recommended fix:

from snowflake.snowpark.modin import plugin
import modin.pandas as pd

ser = pd.Series(["koala", "dog", "chameleon","cat", "mouse", "elephant","lion", "tiger", "bear"])
print(ser.apply(lambda lst: lst[::3]))

Copy