Earlier I indexed Baseline with time_series_column=study_day. This turned out to be excessive. There are hundreds if not thousands of fields; each field has ~700 study days. The UI was very slow to load; searches were very slow; after clicking on a bar the update was slow.
I'm not sure we want to support this many subfields.
One advantage of lower timeout is, if things are completely broken, you find out sooner.
Also update Framingham golden file. I changed the first line of the test to #!/bin/bash -x. Normally I only add the -x for debugging if things are broken, but in this case it's good to always have. Without it, it was hard to tell what failed.
Earlier I indexed Baseline with time_series_column=study_day. This turned out to be excessive. There are hundreds if not thousands of fields; each field has ~700 study days. The UI was very slow to load; searches were very slow; after clicking on a bar the update was slow.
I'm not sure we want to support this many subfields.
One advantage of lower timeout is, if things are completely broken, you find out sooner.
Also update Framingham golden file. I changed the first line of the test to
#!/bin/bash -x. Normally I only add the-xfor debugging if things are broken, but in this case it's good to always have. Without it, it was hard to tell what failed.