Allowing to build plot on smaller energy intervals

[eacharles]

https://jira.slac.stanford.edu/projects/LK/issues/LK-125

[eacharles]

I don't see how the proposed fix really solves the problem. It just offsets it by a decade. The test seems kinda artificial. Would it be better to have a simpler test, like the two number being more that 10% apart or something like that: ie:

if emax / emin < 1.1:
return None
else:
return (emin, emax)

[donhorner]

Copying Description from JIRA:

The method _xrange() of the class AnalysisBase() gives back None,
when emin and emax are too close, in the same decade of values .

I propose to replace the existing line :
if int(num.log10(emin)) != int(num.log10(emax)):
with the following :
if int(num.log10(emin).5) != int(num.log10(emax).5):
...
so for example, we could have plots on short energy intervals using matplotlib,
and exploit the xrange input parameter of MPLPlot() in MPLPlot.py

[donhorner]

Screenshot of JIRA comments:

[jasercion]

Going to roll this into the in-progress plotting overhaul we're working on using matplotlib.