Plot surv by strata

[jburos]

Minor feature added to plot_survival, to support a strata keyword.

At the moment, this results in a separate survival plot or analysis being produced, one for each stratifying group.

E.g.:

(note that I have greyed out the sample names although in reality they printed as text).

[coveralls]

Coverage decreased (-0.4%) to 52.786% when pulling e66084c683b93eac3352382ac1b755ef4f2ce6d6 on plot-surv-by-strata into 744d9b3cef7f5431d77d45b60e6ee7b8fec69c02 on master.

[coveralls]

Coverage decreased (-0.8%) to 52.468% when pulling cea549271199e8d6789db47c5dd73db61786260b on plot-surv-by-strata into 692652307e339826bdd13253efd1e7b208e34b37 on master.

[coveralls]

Coverage decreased (-0.8%) to 52.468% when pulling cea549271199e8d6789db47c5dd73db61786260b on plot-surv-by-strata into 692652307e339826bdd13253efd1e7b208e34b37 on master.

[coveralls]

Coverage decreased (-0.8%) to 52.468% when pulling cea549271199e8d6789db47c5dd73db61786260b on plot-surv-by-strata into 692652307e339826bdd13253efd1e7b208e34b37 on master.

[armish]

@jburos: this is completely off-topic but if you hate those matplotlib error messages as much as I do, just shutdown your jupyters (or anything that makes use of matplotlib) and rm -f ~/.cache/matplotlib/fontList.*, which should get rid of those messages once you re-import the matplotlib once and for all.

[coveralls]

Coverage decreased (-0.9%) to 52.335% when pulling 827753008f352d15aae19ea0fc2c2b95b5d712ea on plot-surv-by-strata into 692652307e339826bdd13253efd1e7b208e34b37 on master.

[jburos]

@tavinathanson same here - neither of these is urgent, but wanted to let you know that the comments were addressed

[coveralls]

Coverage decreased (-0.6%) to 52.689% when pulling f4612fbed04dc8751ff72f6ff6f6fe3678428c40 on plot-surv-by-strata into 692652307e339826bdd13253efd1e7b208e34b37 on master.