Closed rmitsch closed 6 years ago
Implemented as Preprocessing._remove_dataframes_without_annotatation(dataframes)
(in my branch).
@Lumik7 @MoBran Closing as resolved if you agree with the proposed solution.
I already fixed this bug and pushed a solution to the master branch. As I wrote my solution prior to talking to you about invalid recordings, I did not drop the data and initialised the annotations with empty values. This solution can be seen in the get_trip_summaries method.
Anyway, now that we know that this are useless recordings they can be dropped.
I agree with the proposed solution.
Only if chosen token is not KEY_LUKAS. Reason: Mismatch between number of start/end times and number of trips, since some trips apparently don't contain annotation data (e. g. no time or transport mode) and hence are useless.
Proposed solution: Method
_remove_dataframes_without_annotation(dataframes)
dropping all trips w/o annotation data.