df.filter ValueError when going through the Pyam Tutorial

[EmiFej]

I was following the Pyam Tutorial when I encountered the following error:

ValueError: This IamDataFrame is empty!

When entering:

display_df = df.filter(model='XX', variable='YY', region='ZZ')
display_df.timeseries()

When I write variable = 'YY' the error appears, but if I just filter on model and region I get a table of values. Another issue I have with this is that it does not seem to filter correctly. Regardless of what model I chose to filter on, it only provides information about AIM/CGE 2.1. I have looked at the tutorial_data.csv file provided in the repository and there I can clearly see that there is a variable for Primary Energy in a given scenario, yet the error occurs if filtered by that variable.

Any thoughts as to why this error appears?

Lastly, also regarding the df.filter, when typing:
df.filter(region='World').region
this is the output:
['R5ASIA']
When it in fact should be ['World']

[danielhuppmann]

Thank you for reporting this issue! I think that this might be due to an outdated pandas version - can you please check if you have pandas>=1.1 installed? Also, for the future, please report any issues related purely to pyam in the pyam repo or in the pyam Slack workspace #helpdesk channel.

[EmiFej]

Thank you for the swift response, and yes, I should have posted on the pyam repo..
My version of pandas is 1.1.4

[danielhuppmann]

So I created a new pyam branch and ran CI on GitHub Actions specifically using pandas v1.1.4 - all tests pass, so I'm at a bit of a loss for what might be causing your issues.

https://github.com/danielhuppmann/pyam/runs/2950475879?check_suite_focus=true

Can we connect via the pyam Slack workspace?

[danielhuppmann]

Identified the culprit and created a new issue in the pyam repo