InVEST 3.12.0 and 3.14.1 pollination module

Software Support
1

Dear NatCap Alliance

Is it true the adaptations from InVEST 3.12.0 to 3.14.1 are not related to the pollination module?

Meaning that the pollination index output would be the same for both InVEST versions?

Kind regards

Sibylle

2

Hi @sibylles,

There actually were some changes made to the pollination model in 3.14.1:

  • Pollination

    • Fixed a regression where nodata values were not being properly compared. This was only an issue in some development builds after 3.14.0. (#1458)

    • Replaced custom kernel implementation with pygeoprocessing.kernels. Convolution results may be slightly different (more accurate).

A description of the changes made in each InVEST release (including general InVEST changes and changes to the workbench) can be found in our HISTORY notes: https://github.com/natcap/invest/blob/main/HISTORY.rst. I’d also add that a couple bugs in Crop Pollination have been discovered and fixed since 3.14.1, so I would recommend updating to the latest version of InVEST!

3

Thank you very much @csimpson

I am comparing InVEST 3.14.0 with 3.12.0. I cannot use the newest version, because our adapted InVEST python script is based on 3.14.0. For validation I would like to compare the output of our adapted InVEST (based on 3.14.0) with output from another study using standard InVEST 3.14.0.

Of course, if there are significant changes, I will have to recalculate everything for validation. I would like to avoid this if possible.

4

For validation I would like to compare the output of our adapted InVEST (based on 3.14.0) with output from another study using standard InVEST 3.12.0

5

Looking at the history, it does not appear that there were notable changes to Crop Pollination between 3.12.0 and 3.14.0. However again, I would recommend looking over post-3.14.0 changes to the model to check if the mentioned bugs will affect your outputs!