deppen8 / leiap Goto Github PK

When importing artifacts from the database, 10 of them did not have corresponding survey points. This raises a couple of questions:

1. How did these artifacts get entered if there are no corresponding points? Were the points deleted accidentally?
2. What should we do about it?

For now, I have added a .fillna(0) to the Eastings and Northings in the leiap.spatial.find_geo_field() function. This allows the spatial join (gpd.sjoin()) to proceed without throwing an error, but it will cause headaches later if you try to map these artifacts, so it is not ideal behavior.

For the record, these are the artifacts:

• 160060-44-202-001
• 160060-44-202-002
• 160060-44-202-003
• 160060-44-203-001
• 160060-44-203-002
• 160060-44-204-001
• 160060-44-204-002
• 18038b-65-003-001
• 18038b-65-003-002
• 18038b-65-003-003

There are some key decisions to be made still:
1. Use IPy widgets instead of bokeh widgets? IPy widgets CAN control bokeh plots (see bokeh docs), so that seems promising.

2. Given question 1, should we stick with bokeh or try things with altair instead? My inclination is to stick with bokeh because of altair's problems with really big datasets (though this might be a problem with bokeh too).

All survey points have a FieldNumber. Ideally, the Easting and Northing should lie geographically within that field. However, that is not always the case: sometimes field boundaries are too difficult to see, sometimes the GPS is not quite accurate enough near field edges, and sometimes there are simply errors in recording.

We should create/update a new column called geo_field that finds the actual geographic FieldNumber of the point. This will require loading the fields shapefile and using geopandas or something similar to find the intersection.

Not sure if this should be part of the checks.py module or something else.

The bot created this issue to inform you that pyup.io has been set up on this repo.
Once you have closed it, the bot will open pull requests for updates as soon as they are available.

There are some things left to do on report.time_span_chart() to make it look like the one produced with matplotlib for previous annual reports.

• move time values to top axis (or duplicate them there)
• add count values to each production entry
• move time period labels outside the chart area
• add a count of artifacts without time period outside the chart area
• decide whether to use size or color to represent proportion; probably bad practice to use both