Package initial version #3
Conversation
|
@Laura2305 I've managed to get everything to run through on the example data, with some requirements noted down that I had to figure out, but all good. I will add my review by Thursday. |
|
@Laura2305 I would like to propose some code changes in a new branch as part of the review, which you can then merge into setup and we can proceed with the PR. To do this I need write access to the git repo (and also I have some similar changes for the example). Could you give me write access? |
There was a problem hiding this comment.
I've gone through and finished my review of the code. In general it looks great, but I have proposed some changes in my branch #4. Please take a look at my proposed changes, they are mostly related to a function name and adding the REPL dependency explicitly, as well as the filtfilt call with less than 5 slices. These three topics must be addressed, but the formatting changes are up to you if you'd like to keep them or not.
|
|
||
| SCT reference: https://spinalcordtoolbox.com | ||
| """ | ||
| function callSCT(params::Dict{Symbol, Any}) |
There was a problem hiding this comment.
Add a warning or a check for sct installation present, as well as fsleyes
src/Unwrap.jl
Outdated
| padd_size = size(correlation, 1) | ||
| corr_padd = zeros(Float64, padd_size) | ||
| corr_filt = cat(corr_padd .= correlation[1], correlation, corr_padd .= correlation[end], dims=1) | ||
| corr_filt = filtfilt(filter, corr_filt) |
There was a problem hiding this comment.
The call to filtfilt can fail for small numbers of slices, see my PR for a possible workaround or we just check if the sizes are compatible and throw an error if they are not.
There was a problem hiding this comment.
Right, this makes sense. We can just skip that part and use the correlation sign for a small number of slices.
| @@ -0,0 +1,54 @@ | |||
|
|
|||
| function test_AdjustData(datadir::String) | |||
| end | ||
|
|
||
|
|
||
| function testdata(datadir::String, tmpResdir::String) |
There was a problem hiding this comment.
Good testset in general. Maybe we should look at setting up a CI run that checks code coverage
Update CI to use a julia version within the scope of MRINavigator's compat
|
@Laura2305 The failing CI runs are fixed in my branch. I've changed the CI to use Julia 1.3 and 1.8, let me know if this is alright |
1.3 isn't supported by MRIFiles it seems... changing to 1.6
@alexjaffray that's great. Thanks! |
Proposed formatting changes and Function redefinition
fix filtfilt for low slices number
No description provided.