RichardWaiteSTFC

As discussed offline, it seems like there are a couple of differences: 1. There is some discrepancy between the new and old peak positions: qualitatively the new convolution looks off...

I still see spurious peaks at TOF > 10000. If I add a these lines ``` intens_over_sig[~np.isfinite(intens_over_sig)] = 0 uniform_filter(intens_over_sig, size=(3,3,3), mode="nearest") ``` after here https://github.com/mantidproject/mantid/blob/efe46f4f384aa2d37ce26b369dcd0106056fa6cf/Framework/PythonInterface/plugins/algorithms/FindSXPeaksConvolve.py#L167 then it seems to...

Also it looks like some of the commits after my first review could be squashed to make the history a bit cleaner?

Looks great, thanks!

The python prototype code assumes the cut is along the first dimension - in actuality you will need to check which dimension has nbins > 1 e.g. `ws_cut.getDimension(0).geNBins()` produces `50`...

Think this relates to POLARIS scripts - I have the configs files, cal files etc. here if it helps [POLARISPowderDiffraction.zip](https://github.com/mantidproject/mantid/files/14361834/POLARISPowderDiffraction.zip) I have a script (but it's not for producing PDFs...

> I could see where the problems was with issue number 2, but wasn't sure how to recreate. > > For the first problem, when the model is set up,...

Thanks for picking this up @warunawickramasingha if you want to run the algorithm from the python side to familiarise yourself with it - you can follow the instructions here https://github.com/mantidproject/mantid/pull/34762

One suggestion for accepting an event or not relates to determining the angular extent of the collimator opening as seen from the detector. Do you think this would be possible?...

Following discussion with Joe K on ENGIN-X - we will require the users to specify the collimator height and it is OK to center the collimator opening on the sample-detector...