In the filter dialog, columns from the transformed results table are available, suggesting that they can be filtered as well. However, in Manager.process(), the function filter(df) is applied before summarize(df). The problem probably is that users might want to filter at different stages: before the transformation as well as afterwards.
For example, users may want to filter from the results table those rows from a Frequency list that have a frequency below a certain threshold. However, they may also want to exclude rows that match a filter from the frequency list calculation (e.g. exclude matches from a specific subcorpus).
For the time being, the easiest solution is probably to add the option of adding filters either before or after transformations.
In the filter dialog, columns from the transformed results table are available, suggesting that they can be filtered as well. However, in
Manager.process()
, the functionfilter(df)
is applied beforesummarize(df)
. The problem probably is that users might want to filter at different stages: before the transformation as well as afterwards.For example, users may want to filter from the results table those rows from a Frequency list that have a frequency below a certain threshold. However, they may also want to exclude rows that match a filter from the frequency list calculation (e.g. exclude matches from a specific subcorpus).
For the time being, the easiest solution is probably to add the option of adding filters either before or after transformations.
This issue is related to issue #235.