Modified comparison plot, to plot simultaneously T and W tests #132
Conversation
|
cool @pabloitu! let me know when you think its ready to merge. |
csep/utils/plots.py
Outdated
| def _get_marker_w_color(distribution): | ||
| """Returns matplotlib marker style as fmt string""" | ||
|
|
||
| if distribution < 0.05: |
There was a problem hiding this comment.
should we also add the percentile argument for this plot as well?
csep/utils/plots.py
Outdated
| ylims = plot_args.get('ylims', (None, None)) | ||
| capsize = plot_args.get('capsize', 2) | ||
| linewidth = plot_args.get('linewidth', 1) | ||
| markersize = plot_args.get('markersize', 2) | ||
|
|
||
| fig, ax = pyplot.subplots() |
There was a problem hiding this comment.
can you replace this line with the following code? this will let us plot this using a previously initialized axes
if axes is None:
fig, ax = pyplot.subplots(figsize=figsize)
else:
ax = axes
fig = ax.get_figure()There was a problem hiding this comment.
done, and added axes as function argument
There was a problem hiding this comment.
awesome! ill merge this change once the CI is finished
…pyplot.ax() object and percentile for W-test confidence interval
Codecov Report
@@ Coverage Diff @@
## master #132 +/- ##
==========================================
+ Coverage 53.25% 57.06% +3.81%
==========================================
Files 19 19
Lines 3091 3142 +51
Branches 445 454 +9
==========================================
+ Hits 1646 1793 +147
+ Misses 1352 1238 -114
- Partials 93 111 +18
Continue to review full report at Codecov.
|
|
merging now |
Added customization options to plot_comparison_test, similar to those one of Poisson_consistency tests, and added functionality to plot simultaneously a list of w-test if provided. T-test bars are coloured by significance (green is sign. better, red if sign.worse and gray if not significant. W-test significance is shown with a filled marker at the mean IG.
However, this needs an example/tutorial so the "legend" can be explained.