Generating the shapefiles

[alacst]

Hey,

I've been searching for the answer but couldn't find a solution.
How can I get the required shapefiles for my own dataset?

Thank you

[danhamill]

You would have to develop the ground truth polygons yourself. I used GIS software to digitize them.

[alacst]

And what was the base of these? The echograms or dephtmaps?
Thank you

[danhamill]

You would draw your polygons on top of the georeferenced echograms. Something like Figure 4 in my paper: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0194373 Make sense?

[alacst]

so if I don't have visual validation I can not use this software?

[danhamill]

Ground truth data are required to asses if the models are working properly. Riverbeds are dynamic systems and the metrics I found useful for the Colorado River will likely be of limited use at your site.

On March 22, 2021, nnk2193 ***@***.***> wrote: so if I don't have visual validation I can not use this software?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2 (comment)- 804083396>, or unsubscribe <https://github.com/notifications/unsubscribe- auth/ACX5GUW76GBFZC2VOPVOUNTTE5EHZANCNFSM4Y5SG73A>.

[alacst]

thank you! seems like i'm gonna have to find another way

[dbuscombe-usgs]

@alacst You should read the accompanying paper - it would have saved you some time. This is a supervised Machine Learning approach, meaning it is ideally calibrated to each site. You could use Dan's model for the Colorado River in your river, and it may work well, however Dan knows that the Colorado River is somewhat unique from a global perspective, and since we don't know what your intentions are,he's telling you that it may not work and you'll likely have to go through the calibration procedure for your site (fairly trivial if you can identify sediment types with your eyes - if not, then good luck getting anything to work!). He has provided code and detailed in the paper how to do this.

In the future, please submit github issues with more details about hat you are trying to accomplish, especially if you can see the repository has not been worked on for a while (i.e. we've moved on to other things). Good luck!

Best,
Dan

[dbuscombe-usgs]

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0194373