PNDSPY1156¶

Description¶

This issue appears when the SMA detects the use of a pandas element that has a direct equivalent in Snowpark pandas, but some scenarios might behave differently than pandas.

Missing or Unsupported Parameters: limit

Reason: Method nearest is not supported.

Scenario¶

A method with a few scenarios that aren’t supported in Snowpark.

import pandas as pd

df = pd.DataFrame({'A': [1, 2, 3]}, index=pd.date_range('2023-01-01', periods=3, freq='D'))
resampled = df.resample('D')
result = resampled.fillna()

import snowflake.snowpark.modin.pandas as pd

#EWI: PNDSPY1156 => pandas.core.resample.Resampler.fillna has a partial mapping, with few scenarios not supported. Check Snowpark pandas documentation for more detail.
df = pd.DataFrame({'A': [1, 2, 3]}, index=pd.date_range('2023-01-01', periods=3, freq='D'))
resampled = df.resample('D')
result = resampled.fillna()

Recommended fix¶

The parameter limit is not supported in Snowpark pandas. If your code uses this parameter, consider one of these approaches:

1. Remove the parameter: If the parameter is not essential for your use case, simply remove it from the function call.

2. Use default behavior: The function will work with default values for the unsupported parameter.

3. Post-process with native pandas: If the parameter is critical, collect the result using .to_pandas() and apply the operation with native pandas:

   # Convert to native pandas for unsupported parameter
   result = df.to_pandas().fillna(limit=value)
   

   Copy

Behavioral note: Method nearest is not supported.

This behavior may differ from native pandas. Recommended actions:

• Test with a representative sample of your data

• Compare results with native pandas if precision is critical

• Use .to_pandas() if exact pandas behavior is required

Additional recommendations¶

Check the Snowpark pandas documentation to verify which scenarios aren’t supported for that specific element.