stanford-futuredata / inquest Goto Github PK

Hi,

Thank you for your help in the code corrections! I was able to run the code and get the results.

In the experiments, we have 6 video datasets, but what is available here in the code is the proxy scores from the TASTI embeddings, did I get right?
Is there any code providing the scores for the videos and the videos itself?

In line 358, there is

results['median_rmse'] = results_df.compute_median_segment_error(lambda row: compute_median_segment_error, axis=1)

but I could not understand the use of DataFrame.compute_median_segment_error and the lambda function.

Could you please tell me how to figure out this? Thanks so much.

Hello,

When running the current code, two errors appears for each dataset:

The first of them is on the key 'num_segments'

The second one is on the key 'ts'

As a result, the code create the folders for each dataset, but doenst save the result files

The ts problem seems to be just missing the filename on the fuction.

Hello,

Thank you for sharing the repository. I am interested in running the queries and I am stuck in this error when running the command:

SyntaxError: invalid syntax
--- customer-support static ---
File "", line 1
(os.path.join(results_dir, f'results-{config['sampling']['strategy']}-{predicate}-{dataset_name}{special}-alpha-{alpha}-segments-{num_segments}{pilot_str}.pq'))
^

This same error keeps going for all datasets

Any clue what could it be?

Thank you!