Closed ChristopherTracey closed 7 years ago
Step 1 of this is convert the percentage of the PUs to Area (ie. multiply by 10). We used to use area in an earlier version of the test data, but then we switched to percentage as it was available in all the compiled datasets. The current code is summing up all the percents, which is creating some seriously huge numbers. This has been converted back to acres.
Otherwise, this is complete, but still needs testing. The key was to split the list into two dataframes, do the compare between the medium and low, and then join them back up (removing any species from the med/low DF that are present in the High DF.
Included in #72
Currently, the probability of occurrence for a particular SGCN in the AOI is defined by the highest summed proportion of coverage across the three probabilities (high, medium, and low).
Proposed change to have High probabilities, since they tend to represent confirmed occurrences or observations, to override Medium or Low values, even if the proportion of a high probability is lower than either of the two categories. Medium and Low probs will then fight it out based on higher proportion.
This should give more weight to the confirmed occurrences for the AIS.