Closed voisey closed 4 years ago
The reason only certain plots have been affected is because some are tagged with @figuregen while some are tagged with @figure. @Zaharid, would it be sensible to edit savefig so that the figures tagged with @figure are put in columns too?
@scarrazza @Zaharid I wasn't totally clear on the outcome of yesterday's discussion on this. Is this ready to be reviewed/merged or is there something else you'd like to be done?
@voisey, fine by me from a conceptual point of view.
I think we should merge this more or less as is.
A few more general points:
It would probably better to add a class
instead of a style
tag to these reports. And then set the set the style for the class in the css file.
@voisey probably we could look into this https://primer.style/css/objects/grid#inline-block-grids we are already using this css style so adding the right classes should do the trick.
On your last point @Zaharid, do you mean that we can ditch the float business altogether and use d-inline-block
instead? I don't really understand how this would work practically. Should something like:
<p class="d-inline-block col-2 p-2">
<img class="width-full avatar" src="..." alt="..." />
</p>
work in our case?
Either way, I've put the styling I used in a CSS class
Regarding your first point in https://github.com/NNPDF/reportengine/pull/33#issuecomment-618404402, where did this file come from?
@voisey there is this https://github.com/NNPDF/reportengine/tree/master/src/reportengine/styles the CSS style is some old version of Primer CSS that @scarrazza got. The template is a pandoc template minimally adapted to work with the style.
Can we merge this and then open an issue about updating the styles?
Ok I'm going to stop trying to "improve" this. It is already an improvement. Thanks!
Haha, sure. You're welcome!
Works towards closing https://github.com/NNPDF/nnpdf/issues/268. So far it works for the 'PDF plots' section of the vp-comparefits report (https://vp.nnpdf.science/KbFpz1hOQwKsPbFVaDtqew==/pdf_report_report.html), i.e. for the PDF plots and the effective exponents. Personally I think it's much more readable with this formatting and the PDF plots page being shorter is certainly nicer. I think the same formatting should be applied to the other plots in the report too: namely distances, arc lengths, training-validation splits, and positivity plots. Let me know if you have any thoughts