Kilian Vos

hi @tonycoppola , Did you use the same image (same .tif file) for your analysis in QGIS? The NDVI is good because, as you mentioned, both bands come at the...

hi @sac3g15 , I agree, georeferencing errors are surely an issue when using the medium resolution imagery. Landsat products provide a georeferencing root-mean-squarred-error with each image, which can be used...

I hope that the new release of CoastSat (v2.0) should fix this issue of the bias so I will close this issue for the time being.

hi @tolandwehr , have you tried using the 'points' option when saving the shorelines as GeoJSON? The problem when using Linesterings is that the toolbox works with numpy arrays, which...

hi @Lvulis , make sure yo ustart in a new data folder so that you are not mixing up downloads from the previous and new versions of coastsat. I'm also...

ok I also ran into that corrupted zip file error, tried to open it on windows and it said it was unable to unzip it. It's probably a connection timeout...

I just pushed a commit on the `development` branch with a try/except loop around the download functions, it tries to download it up to 100 times to overcome timeouts and...

great, thanks for testing. Will keep it like this for the moment, if it still occurs we can re-evaluate an alternative solution

hi kat, I trained the classifier with different training datasets, for example the 'bright' sand, I digitsed pixels on beaches with extremely white sand. You can create training data and...

hi Kat, 1. The training data used for the default classifier, both L8 and S2 is contained in the github repo in these pickle files https://github.com/kvos/CoastSat/tree/master/classification/training_data. I think it's the...