Thank you very much for your answer.
In fact you totally get the point, I proceed the way you showed in the example you posted.
The only difference with my real life :
– I have ~2 years of dates multiplied by ~100 categories = more than 50.000 bins, and DataGraph show me an orange dot on the Pivot section, telling me that there are too many bins. The extracted data seems to be consistent, still.
– I click on hide in the Pivot command and the graph is entirely plotted by the Plots command, but with correct axes
In my real file, categories are regions and numbers are hospitalisations or deaths (Covid data), and I keep in the data I extract from my Pivot the “day” granularity (no groups by months or weeks)
As the Pivot command graph is for me “useless”, I was thinking of generating the Pivot table without the Pivot command, but perhaps with Plot Action before graphing with Plots Command. In this case, I choose the “bin” action (pdf manual P.28, screenshot below), but my categories are in red as it seems I can only select numbers and not text.
Arnaud