Skip to content

davidpng/Myeloid_Machine_Learning

Repository files navigation

README

AUTHOR: David Ng, MD University of Washington Department of Laboratory Medicine DATE: 17 September 2014

/Input_Outputs/Comp_libs This folder contains spectral overlap coefficents for the various flow tubes - T-cell tube not collected yet

These are python classes accessed through Main.py that will:

1 find FCS files according to a case list and file structure (HEADER_FIND_FCS_files.py) 2 Read in uncompensated FCS files (HEADER_Import_FCS.py) 3 Bins FCS files and outputs cases (sets of FCS files) as a sparse row matrix (HEADER_ND_Binning.py) 4 Saves these results (and metadata) to an HDF5 file as follows (HEADER_Generate_HDF5.py):

''' /tube//tube1 tube2 tube3
    /data/<case num>/data
                    /indices
                    /indptr
                    /shape
n.b.: The saved elements in data are sufficent to regenerated a scipy.sparse.csr_matrix data stucture

'''

HEADER_Read_FCS.py Defines a class that handles reading of the HDF5 file. Work in progress

About

IO handling and process of raw FCS files

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages