Leon Lampret

If I run that jupyter cell several times, over different days, I always get the slightly negative result with `read_sql_query`. According to DBeaver, the min and max values in my...

Of course, if I run `! pip freeze` in a jupyter notebook in SageMaker Studio on AWS, I get: ``` alembic==1.10.3 appdirs @ file:///home/conda/feedstock_root/build_artifacts/appdirs_1603108395799/work asn1crypto==1.5.1 asttokens==2.2.1 attrs==22.2.0 awscli==1.27.70 awsio @...

@kukushking Have you been able to reproduce the bug? If not, shall I post here a dataframe containing the table that causes the issue?

@kukushking I was not able to reproduce the wrong output by uploading a dataframe, but I have found the source of the problem: internal casting of redshift's `numeric(38,20)` to pandas's...

@etiennedeg Thank you for your suggestion. The documentation for this function is very sparse: ``` help?> GraphsMatching.maximum_weight_maximal_matching_hungarian No documentation found. GraphsMatching.maximum_weight_maximal_matching_hungarian is a Function. # 2 methods for generic function...

I think this package has great potential, since optimization algorithms are often used in industry (3 years ago, I succeeded in a job interview because I used scipy's `min_weight_full_bipartite_matching` to...