Thanks, do you maybe want to update the example, too? Also, could you please redirect your PR to the development branch?

On it.
Incidentally, don't we want to have a shorthand for sklearn, i.e. instead of doing

flow = flows.sklearn_to_flow(clf)
run = runs.run_flow_on_task(task, flow)

do something like

run = runs.run_sklearn_on_task(clf, task)

From what I've seen, people always use those two lines together without doing anything with the flow object.

We have run_model_on_task.

I updated the example, but there was nothing really wrong with it?

@joaquinvanschoren I did some more changes - could you please also have a look again and approve it if you like it?

Hmm, Travis fails because it can't find numpy in the Python 3.4 test. What should I do about this?

Codecov Report

Merging #456 into develop will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff            @@
##           develop     #456   +/-   ##
========================================
  Coverage    89.46%   89.46%           
========================================
  Files           32       32           
  Lines         2771     2771           
========================================
  Hits          2479     2479           
  Misses         292      292

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4603345...d24c04c. Read the comment docs.

Oh oh, it seemed like Travis finished successfully, but then suddenly there was still a fail after I merged it.
The issue is that Travis fails because it can't find numpy in the Python 3.4 test. Any ideas?