InVEST 3.8.0 released today

Today we released InVEST 3.8.0. It has been 9 months since our last official release! That means there’s lots of good stuff in this one and we strongly encourage all users to update! Here’s a glimpse:

  • Two new InVEST models: Urban Flood Risk Mitigation & Urban Cooling

  • natcap.invest is fully upgraded to Python 3.7 and Python 2.7 is no longer supported

  • Much improved validation of input data for all InVEST models

  • An improved command-line-interface to facilitate batch-processing, or for those who just prefer to do it from the terminal.

  • Updated sample datasets for many models

  • A rebuild of DelineateIt which is faster and uses Taskgraph to avoid re-computation

  • Important bugfixes and speedups for Coastal Vulnerability (all users should upgrade!)

  • A fix for a critical issue where SDR would hang or crash on MacOS

  • Many more bugfixes across models including SDR, Pollination, Recreation, Coastal Blue Carbon, Habitat Quality, Wave Energy, Seasonal Water Yield, NDR, HRA

  • For a full rundown, see our changelog (https://invest.readthedocs.io/en/latest/changelog.html)

Download the latest version for your operating system at:
https://naturalcapitalproject.stanford.edu/software/invest

Or python users, download the latest from https://pypi.org/project/natcap.invest/

Many of the improvements in this release are a direct result of conversations and bug reports that have happened on this forum! So thank you for all the support and feedback.

The NatCap Software Team

6 Likes

I really would like to do the training course on the InVEST software. But the enrollment is closed…

Hi @NataliaPeriotto, welcome to the forums!

Yes, unfortunately the Stanford-hosted MOOC platform we were using is being shut down by the university, and so enrollment has been closed. We’re still figuring out what to do with the courses, but I believe we’re still planning on offering them somewhere. As for when the MOOC might be available again, I can’t really say other than that we’re pretty occupied with preparing for the Natural Capital Symposium at the moment and I bet the MOOC is something we’ll be dedicating some time to after the symposium.

In the meantime, our InVEST User’s Guide might not have the same guided structure, but it is always available online and might be helpful to you: http://releases.naturalcapitalproject.org/invest-userguide/latest/

I’m very sorry for the inconvenience, and I hope this helps!
James

Hi, James! Thank you very much for your reply!

1 Like

Hi @NataliaPeriotto -

We are working on an updated version of the MOOC. I can’t promise when it will be made available, since we’ve had several delays already, but we will let everyone know on the forum when it is.

~ Stacie

Great to know this!

I have recently run the Coastal Vulnerability model with one of the latest version (November 2019), that Dave particularly sent to me in this forum (InVEST_3.7.0.post576+h92e979470c93_x86_Setup.exe) and I would like to know which are the new improvements in 3.8 with respect the one I used?

Thank you very much for your help and attention!

Cari

Hi @Cari, since that post576 build, it looks like there were a couple small improvements and fixes that may or may not apply to your runs and shouldn’t affect the results at all.

This issue was addressed: https://bitbucket.org/natcap/invest/issues/3934/cv-surge-exposure-should-not-depend-on-aoi

And better validation of input data in the user interface was implemented.

I certainly recommend upgrading. But if your question is, “is it necessary to re-create all my results with the new version?” Probably not.

Thank you Dave for your help!

I would like to test the different results obtained changing some of the parameters (such as e.g. protection distance, maximum fetch distance, relief radius) and I wonder if the best way to run different times the model using the same workspace is just adding a new “Results suffix” every time.
Would you recommend this?

Thank you very much again for your assistance!

Cari

Yes, that’s a great approach for doing multiple runs. When using the same workspace, the model will be able to re-use intermediate results (by copying files and appending the new results suffix) that already exist and only re-compute the parts that will change.