Closed SFuller4 closed 1 year ago
Custom NaN Values: Strings potentially found in the dataset, intended to be read as NaN values before being sent to Neo4j. If there are no strings supplied, empty cells will be replaced with the string 'null'. Only a list of strings that will apply to the entire dataset is supported through this plugin.
Add default strings: If selected, the default strings that pandas uses will be added to the list of custom NaN values. You can find information about what those default strings are here Pandas.to_table() in the na_values section.
@StanislasGuinel I updated the request and resolved your comments. I've also added a comment that has the data for the documentation page.
I also added the new parameters to the pytest file, although I was unable to perform the pytest locally.
Let me know if there are any other problems so I can help get this in place as soon as possible!
Thanks for the comment for the documentation page !
Did you test that your feature work ? It still fails (TypeError: create_dataframe_iterator() got an unexpected keyword argument 'na_value'
)
Also, there are many line diffs in this PR that shouldn't be there, I guess you used a formatter, but that makes the code review much harder. Could you revert all the useless diffs so that the review can only focus on the new feature code ?
Sorry about that failure, I did run it on our instance, but after your message and checking, I had file discrepancies between what was on the instance and my local machine.
I've also reverted most of the black formatting changes. I was worried about that on the initial pull request, but figured it would be okay since it's generally required for open-source code to be formatted black or lint before pushing.
Those changes are good with me! I went ahead and updated my previous comment for the documentation to use NaN instead of NULL as well, to stay consistent.
Brought the Export Relationships json in line with Export Nodes json.
@SFuller4 the new version of the plugin has been published !
Thanks for the help @StanislasGuinel ! Everything updated and looks great from our side!
This is directly solving this issue [https://github.com/dataiku/dss-plugin-neo4j/issues/29](Issue #29).