search

padsley / k600analyser

Code for the K600 analyser including plugin codes for silicon, clover and HAGAR data
1 stars 4 forks source link

Data classes with multiple detectors #76

Closed padsley closed 8 years ago

padsley commented 9 years ago

I'm an idiot.

padsley commented 9 years ago

More detail: @pellegri says that I'm an idiot.

pellegri commented 9 years ago

I've just find a way to point it out :P

padsley commented 9 years ago

The problem is that the output class pointer (e.g. gammy) is just set to the last working version of the thing that has been filled. This wrong and evil and bad.

It should be set so that the same GammaData (for example) object is used throughout and different sorting algorithms just add to that pre-existing class.

This will require some (hopefully not too major) rewrite of the sorting algorithm.

In future, the pointer to the Gamma/SiliconData class should be handed to the sort algorithm (which should be a void not a pointer to a Gamma/SiliconData class) so that it can just add to the bottom of something that's already there rather than resetting it.

padsley commented 9 years ago

Also, ow.

padsley commented 9 years ago

This change will be made in the SetupGammaDev branch and then transported across into the master when possible.

padsley commented 9 years ago

I've done this for the GammaData classes in the SetupGammaDev branch.

When @pellegri has finished her development work on that branch and we merge it back into the master, the change will make it to the master.

I will make the same change to the SiliconData classes in the master branch at some point fairly soon. It's not urgent as no one requires it (yet).

padsley commented 8 years ago

This was done at the same time as the SetupGammaDev merge back into the master branch, and was expanded to include the silicons.