search

planetirf / math456_soil_project

0 stars 0 forks source link

Fix Depths #6

Closed JessicaShippen closed 4 years ago

JessicaShippen commented 4 years ago

Make a consistent depth system.

planetirf commented 4 years ago

The correct depth classes should be:

0-12 12-24 24-36 36-48 (if it exists)

0-8 is also valid but did not follow the correct sampling scheme.

recursory commented 4 years ago

This issue was at least partially resolved by refactoring depth-1 into depth-low and depth-high, right? If something else needs to be done, could we close this and make it a separate issue?

planetirf commented 4 years ago

The mis-matching depths issue was solved by using the new dataset PCP-R-ready.csv, that was sent from an old student I reached out to last week.

The total observations were reduced from ~1200 to ~850 in the new data set. It also applied a sampling date and an irrigation variable which was not available in the previous set.

The splitting of the columns from a single factor to a top and bottom is sort of a data quality practice and the correct format for the soil-data R packages i'm learning which might be beyond the scope of this class.

The script runs well on this data set and there are no mis-labled values in the new dataset.

[CLOSED]