search

davidssmith / DCEMRI.jl

DCE MRI analysis in Julia
Other
37 stars 16 forks source link

Fix residual comparison for ExtTofts model & Combine plot functions #39

Closed notZaki closed 7 years ago

notZaki commented 7 years ago

Two changes:

  1. In science.jl. Fixed bug with ExtTofts model. Large regions of the QIBA4 phantom couldn't make it past the residual comparison, so the modelmap and residual map had big patches of zeros. However, the parameter maps still got produced correctly because they were being accepted before the residual comparison even took place - an incorrect behavior providing the correct result.

  2. In validate.jl. I combined analyze4() and analyze6() into a single function because they share a lot of code. This doesn't exactly have any definite advantage, but it does considerably reduce the lines of code, and might make future edits easier. The drawback is that the range and ticks on the colorbars now use default values instead of the pre-defined phantom-specific values they had before.
    Since the code for drawing the maps is similar, it could be wrapped in a function to further reduce the amount of code lines - but these all seem to be incidental changes.

davidssmith commented 7 years ago

I like this a lot. Currently trying to figure out why the builds are failing.