Pfthb

I received a python script from an archive from a former employee. It states that I need:

• python 2.7


• numpy 1.92


• scipy 0.15.1


• scikit-image 0.11.3


• scikit-learn 0.16.1


• tifffile 0.5


• pandas 0.16.2


• matplotlib 1.4.3


• pylab 1.9.2

I built an UbuntuMATE virtual machine. This comes with python2.7.15. I then moved on to installing the packages using pip. This went pretty well up to the point where
matplotlib 1.4.3 had a bad "egg". Word on the street was that it was a bad install. I then installed matplotlib 1.5.3.

But I couldn't install pylab. (Could not find version...)

Advice was that with sufficient matplotlib and scipy, I shouldn't need pylab.

I ran the old python script and got lots of errors with matplotlibrc.

As an alternative I tried to install matplotlib 2.0.0. This results in still more complaints from pip, including: unsupport LaTEX, deprecated functions.

Is there a way out of this mess?

Why might my former colleague have indicated pylab 1.9.2 if it cannot be installed?

Why is matplotlib 1.4.3 uninstallable?

Should I wipe out my VM and start over or can I salvage the installation?

pylab is part of matplotlib. It is essentially a small file that makes numpy, matplotlib.pyplot and some matplotlib.cbook and matplotlib.dates functions available under the same namespace.

You cannot install pylab on its own. And there is no pylab-version.

You might notice though that the list you got states numpy 1.92 and pylab 1.9.2. Since (a) There hasn't ever been a version 1.92 of numpy, and (b) there hasn't ever been a version 1.9.2 of matplotlib, what this probably means is

numpy 1.9.2

There are no details given as to why the installation fails. Therefore just note that I was able to install all versions, except tifffile, perfectly fine via conda(-forge)

>conda create -n sometest python=2.7 numpy=1.9.2 scipy=0.15.1 scikit-image=0.11.3 scikit-learn=0.16.1 pandas=0.16.2 matplotlib=1.4.3