[feat] preprocessing helper

[ngupta23]

Description

Based on various user feedback

Although there are tutorials on how to process the data, it would be beneficial to provide helper functions to check or preprocess the data in the right format

Some examples indicated by users

• Flag duplicate rows for combination of id_col, time_col - Could indicate a need for aggregation
• Flag duplicate rows for combination of id_col, time_col and target_col - Could indicate duplicates
• Flag missing dates for each unique_id. Could also indicate to the user that they can use fill_gaps function from utilsforecast
• Flag non-numeric columns (categorical, etc.) in the data
• Flags for short time series (not sure how this will work, need to think about it) - remove fine tuning for short time series, separate data into short and long time series and make 2 calls - short without preprocessing, long with preprocessing.

cc: @cchallu

Link

No response