AndrewJSchoen
commented
8 years ago Definitely. This is an absolute must. The tricky thing will be getting some datasets that don't have participant data in them. Perhaps some non-human phantom data would work well.
On Jul 22, 2016, at 4:09 PM, Andy Pohl notifications@github.com wrote:
I'm just putting this here as a reminder that once we're at a 1.0, i.e. the whole workflow is running, we should probably build up a test suite with toy (small) data, so if we make future changes, we'll see if something unexpected breaks. A test suite could be very helpful in this stage too, but the main priority now is still building up the workflow piece by piece. Test suites can be produced to:
Check individual scripts. Make sure changes to 3rd-party BI software don't affect DIPA/Perseus. Make sure Pegasus changes don't affect anything. I'll try to find some good examples and in particular try to put the pegasus part in place.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.
I'm just putting this here as a reminder that once we're at a 1.0, i.e. the whole workflow is running, we should probably build up a test suite with toy (small) data, so if we make future changes, we'll see if something unexpected breaks. A test suite could be very helpful in this stage too, but the main priority now is still building up the workflow piece by piece. Test suites can be produced to:
I'll try to find some good examples and in particular try to put the pegasus part in place.