Den 2017-01-23 skrev Albert-Jan Roskam <sjeik_ap...@hotmail.com>: > (sorry for top-posting) > I does not appear to be possible in matplolibrc (1). But you can > use matplotlib.cm.register_cmap to register new cmaps (2) such > as these (3). > > (Note: I did not try this) > > (1)http://matplotlib.org/1.4.0/users/customizing.html > (2)http://matplotlib.org/api/cm_api.html > (3)https://github.com/BIDS/colormap/blob/master/colormaps.py
More than a year later I think I can advice on the best workaround: Use virtualenv and you can work with newer versions of Python packages than those available in your Linux distro of base Python installation. I am currently learning to use jupyter notebook, pandas and other nice packages in much more recent versions than available in my Linux distro. I hope I can copy this onto my computer at work -- running another Linux distro where virtualenv3 does not seem to be available. /Martin -- https://mail.python.org/mailman/listinfo/python-list