Updates on progress to cleaning up the project

[Scott Chamberlin]

 

Quick update here.  I’ve gotten the notebooks to a relatively clean point and am in the progress of writing both a tutorial as well as a quick start guide.  Most of the progress is still contained in this branch:

https://github.com/scottcha/OpenAvalancheProject/tree/clean-notebooks

 

As of now you might be able to make your way via two paths:

1. If you want to generate new model data from the starting point of the NCAR .25 degree GFS model you now can just follow the notebooks in the Scratch Notebooks folder in numbered order 1-4 (you need to obtain the input files from here https://rda.ucar.edu/datasets/ds084.1/).  The labels for three regions Washington, Colorado and Utah are located here: https://oapstorageprod.blob.core.windows.net/oap-training-data/Data/V1.1FeaturesWithLabels20131201To20180430.zip
2. I have also provided a quick start to just the ML notebook (notebook 4) where you can obtain the files here https://oapstorageprod.blob.core.windows.net/oap-training-data/Data/MLDataWashington.zip  Note that this process is a quickstart to just the ML portion but currently the default deep learning models quickly overfit with this input.  Once I get all these cleaned up and the tutorials published I’ll be looking in to that next.

 

I understand its still a bit of a hurdle to get up to speed.  Closing that gap is still my top priority.  Unfortunately I moved last week and am still getting settled so progress has been slower than anticipated.

 

Thanks!

Scott