To easily install pyMCR into non-ascii environment (e.g. Windows 10 in Japanese), it would be nice to replace the non-breaking hyphen (U+2011) in line 168 of README.rst into more common hyphen-minus (U+002D) to avoid UnicodeDecordError.

On Demo.ipynb, I guess several mcrals._c_mrd should be mcrals._c_mrd[-1] .

Anyway, this code is what I look for. Thank you so much to share your code!