remove file extension#38
Conversation
SFJohnson24
commented
May 18, 2026
SFJohnson24
commented
- updates all results.csv for non-usdm data
- updates the results script that produces the csv for nonusdm data
| header = ["Dataset", "Record", "Variable", "Value"] | ||
| rows = [] | ||
| for issue in issue_details: | ||
| dataset = issue.get("dataset", "") |
There was a problem hiding this comment.
Where is the csv coming from in the first place? The data in sharepoint doesn't have csv extensions...
There was a problem hiding this comment.
convert() on line 58 in this file--takes the json results output, converts to csv
There was a problem hiding this comment.
instead of adding logic here, can you update the logic at the source that converts these from ae.xpt to ae.csv?
gerrycampion
left a comment
gerrycampion
left a comment
There was a problem hiding this comment.
I also looked at some of the usdm files and I see the path value is messed up. For example, see this file:
https://github.com/cdisc-org/cdisc-open-rules/blob/main/Published/CORE-000940/negative/01/results/results.csv
gerrycampion
left a comment
gerrycampion
left a comment
There was a problem hiding this comment.
it looks like at one point we were using 0-indexing for results, but now we are using 1-indexing. this is throwing off a bunch of the open rules test results.
We don't know which ones have 0-indexing and which ones have 1-indexing, so I'm not sure the best way to fix this. 2 ideas:
- compare the row numbers between sharepoint results and the engine results and replace where all numbers are off by 1.
- figure out the date at which we made the switch and use the file timestamps.
see: #40 for USDM fix |
see #41 for index 0 data that was scrubbed |
gerrycampion
left a comment
gerrycampion
left a comment
There was a problem hiding this comment.
Has merge conflicts, but otherwise good
2 participants
