Convert SPARQLWrapper.SmartWrapper to pandas dataframe

[appliedgraphs]

I want to convert a SmartWrapper to a pandas DF. We have two relevant issues:

• #125 Which is focused on a SPARQLResult
• #1179 - which also seems to be about SPARQLRest

My goal is to execute a SPARQL query, get a result, and then transform to a dataframe. I am certain that I am missing something because this must be a very common need.

[nicholascar]

@appliedgraphs have you seen this existing script: https://github.com/RDFLib/sparqlwrapper/blob/master/SPARQLWrapper/sparql_dataframe.py

[afk314]

I hadn't, but it's nice and easy to follow. It also works wonderfully! What can be done to make this easier to find or otherwise more integrated with rdflib?

[nicholascar]

Yes, I think PANDAS is well known enough to be supported as a standard output option. The only hitch is that we want to ensure basic installation of SPARQLWrapper doesn't require PANDAS, since it's large, so it can be included but a pip install optional will need to be made to enable it.

Something like pandas[pandas] in requirements.txt and then pip install sparqlwrapper[pandas] to install.

Would love to see a PR for this...

[eggplants]

@nicholascar Already:

sparqlwrapper/setup.cfg

Lines 53 to 54 in 2a6e2d3

	pandas =
	pandas>=1.3.5

[nicholascar]

Great, thanks @eggplants, so the installation tasks are already done. So, I suppose, a deeper integration of PANDAS could be made, as long as appropriate warnings are thrown if users try and call PANDAS exports without the necessary installation.

[WolfgangFahl]

this is what pyLodStorage was motivated by https://github.com/WolfgangFahl/pyLoDStorage where you get list of dicts that you can immediately use in pandas see #160