crotwell / pickax Goto Github PK

Sc_quakes.py is currently outputting pick files to a directory on our local machines. If Chase, Logan, and I are all going to be running this program locally, how will we eventually aggregate the resulting datasets?

Here is one idea: Does the department have a dedicated seismology server for which we could be granted write-access (to a specific directory only)? If so, perhaps sc_quakes.py could be modified to output our picks there at the end of each session. Further, if that process were in place, those files could also act as an input to the sc_quakes.py Import step. Instead of sc_quakes.py reloading data (from web services) and then appending our picks to the imported data, it could pull all of our picks from the server. The data on the server would essentially become the latest and greatest source of work to date.

Could we add a filter that lets us look at quakes we have not picked yet?

Request for Enhancement: This functionality does not necessarily have to occur in real-time from the same interactive window. Perhaps there is a different window that would be launched by the script if we wanted to work in 'edit mode'.

Request for Enhancement: If we accidentally hit 'a' instead of 's', it would be nice to correct that mistake in real-time, within the interactive window.

Request for Enhancement: As we learn how to recognize p- and s-waves, it might be helpful to hide the USGS picks in the beginning. This could just be a filter setting in the interactive window so we can toggle back and forth.

We talked about possibly making all of the amplitudes on the same scale.

Fix it!

Ashley's Error: After completing some picks this morning, I noticed these comments in the log. The message seems to be inferring that the picks I made while using one of the filters will 'not be a valid QuakeML file'.

"UserWarning: 'bandpass 1 10' is not a valid QuakeML URI. It will be in the final file but note that the file will not be a valid QuakeML file."

Phillip's Response: The issue is that I put in the filter_id in the QuakeML Arrival object as just whatever the name of the filter is. However, QuakeML has a rather strict opinion of the structure of an "id"...I should make a legit filter_id eventually, but just ignore until then."