Besides missing documentation (e.g.: #12), it is not clear how and where to update the documentation of modules, e.g. after renaming or deleting modules (for list see #13 ).
In addition many of the bugs, discovered in testpipelines refer to a few implicit assumptions of original CP, which were not considered / formally tested by developers. It would be good to somewhere have a short manual for developers.
Specifically these assumptions are important:
At least one pixel is background after segmenting primary objects.
Background is defined by 0.
ObjectIDs are continuous. They start with 1 (thus the amount of objects corresponds to the maximal ObjectID)
single plane microscopic images are imported into CP as double(OriginalRawIntensities) ./ (2^16 -1)
if no object is present, the saved measurement should be a single line of 0s , where the amount of columns corresponds to the different features; e.g: centroid location of a site without an object should be [0 0] ; where the two columns correspond to the X and Y coordinate
Besides missing documentation (e.g.: #12), it is not clear how and where to update the documentation of modules, e.g. after renaming or deleting modules (for list see #13 ).
In addition many of the bugs, discovered in testpipelines refer to a few implicit assumptions of original CP, which were not considered / formally tested by developers. It would be good to somewhere have a short manual for developers.
Specifically these assumptions are important: