It might be a good idea to call update() when calling the install CLI function. This could be done automatically each time (to update the index before downloading). An alternative would be to do it the first time install is called. The latter could be done by catching an error in line 53 of cli.py which can occur if the user did not call update before calling install for the first time and therefore the pdfsets.index file does not exist.

The above would remove the need to run python3 -m parton update before installing a pdf set for the first time (or possibly ever).

Alternatively, an error could be caught in line 50 of io.py, which is raised if pdfsets.index does not exist, to remind the user that update should be called first.

I'm trying to use the mkPDF function with PDF sets I already have downloaded. I believe I'm properly using the pdfdir argument, since it's able to find the .info file I'm trying to load. However there is an error with the load function (shown in the attached photo) that seems to be internal to the parton package.

The error is “ load() missing 1 required positional argument: ‘Loader’ “

Is this an internal issue or am I doing something else wrong?

Thanks,
Steven

To get parton luminosity ratios matching https://www.hep.ph.ic.ac.uk/~wstirlin/plots/plots.html, I had to multiply a factor of 1/s into the output of PLumi.L().

For example, this plot (with the 1/s factor included) matches lhclumi7813_2013_v1.pdf:

whereas without the 1/s factor, the result does not agree:

Perhap this is a misunderstanding of the exact quantity that PLumi.L() returns; it would be good to clarify the definition and units in the README.