update make_data_column_uniform #125
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dakshvar22
requested review from
dakshvar22
and removed request for
janaab11
dakshvar22
reviewed
Feb 24, 2023
slu/slu/utils/preprocessing.py
Outdated
| @@ -74,9 +74,11 @@ def make_data_column_uniform(data_frame: pd.DataFrame) -> pd.DataFrame: | |||
| ): | |||
| if isinstance(row[const.ALTERNATIVES], str): | |||
| data = json.loads(row[const.ALTERNATIVES]) | |||
| if isinstance(data, str): | |||
There was a problem hiding this comment.
Why would json.loads yield a string? Shouldn't it strictly give a dict?
There was a problem hiding this comment.
Production calls in vodafone contain noisy transcripts, where ASR alternatives seem to be dumped twice. The second if condition is supposed to fix this by performing another json.loads(data).
Case example:
slu/slu/utils/preprocessing.py
Outdated
| @@ -58,7 +58,7 @@ def make_reftime_column_uniform(data_frame: pd.DataFrame) -> pd.DataFrame: | |||
|
|
|||
| return data_frame | |||
|
|
|||
| def make_data_column_uniform(data_frame: pd.DataFrame) -> pd.DataFrame: | |||
| def make_data_column_uniform(data_frame: pd.DataFrame) -> None: | |||
There was a problem hiding this comment.
Why does the type hint need a change?
There was a problem hiding this comment.
This was a mistake, return type hint shouldn't be removed. Fixed this in a new commit.
Update return-type hint
dakshvar22
reviewed
Apr 20, 2023
Comment on lines
+84
to
+87
| data_frame.loc[i, const.ALTERNATIVES] = json.dumps( | ||
| data, ensure_ascii=False | ||
| ) |
There was a problem hiding this comment.
discard this data point.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Ticket: https://vernacular-ai.atlassian.net/browse/CMS-2525
Data path: s3://vodafone-depository/datasets/problematic_datapoints_(production calls).csv)
Problem:
I found that a large amount of production data in Vodafone contains noisy transcripts.
In these data points, ASR alternatives seemed be dumped twice (double
json.dumps()).The
merge_asr_outputplugin values are incorrect for these datapoints, which eventually passes as input to the model.Following is an example of a noisy transcript, and its expected vs actual
merge_asr_outputvalue.Sample transcript:
Expected output:
Actual output: