Hello:

It looks like the only difference (please correct me if I'm wrong) from FCSTrans implementation is that flowCore uses it's own C++ implementation of biexponential_transform() in place of original C implementation in FCSTrans.

1) If FCSTransTransform() was meant to reproduce FCSTrans, why wasn't the original biexponential_transform() used?

2) Is it possible to find the original C source? I looked on Sourceforge:

https://sourceforge.net/projects/immportflock/

but it's not found there.

Regards, Nik Tuzov

Well, this was a bit before my arrival on the team for flowCore, but I think most of the answers to your questions can be found in this commit:

https://github.com/RGLab/flowCore/commit/30d9fc3fd7e46f9b75a45c6410b4a9c43233a8d8

The original C implementation of biexponential_transform is there, along with its C++ replacement.

1) I believe Mike was just updating the legacy C code to hook it in to Rcpp::compileAttributes for auto-generating the R bindings. See other commits from right around that time: https://github.com/RGLab/flowCore/commit/a5aaaf39466699767b14ad25a6e93ede600cb540 https://github.com/RGLab/flowCore/commit/df4d7ef382b46761920a34fbda8a43e8aad939b4

2) Agreed, the sourceforge link (including yours SamGG), has the R source that calls biexponential_transform but not the C source for biexponential_transform. You can find that in the first commit listed.

Mike may have more insight, however.

Hope that helps, -Jake