Hi Nick,

As far as I know MGET is compatible with Windows 8.1, but I have not tested it. We do not have access to anything beyond Windows 7 at this time.

MGET is currently compatible with ArcGIS up through version 10.2 but not 10.2.1 yet. I have a workaround to enable it on 10.2.1 and anticipate releasing a compatible version sometime this month.

In general, ArcGIS users who wish to use MGET install ArcGIS first and allow it to install Python, before attempting to install MGET. Please see here for detailed installation instructions. If you can give me some more details about the error—what exactly it said and when it happened—I might be able to help further.

Your tagging project sounds interesting. Can you tell me a bit more about it? I suspect your question is ultimately statistical—what statistical approach is appropriate for your data. I might be able to advise if I knew more.

From the practical side, assuming you have a collection of points that have dates, MGET can help you obtain environmental values from popular remote sensing and ocean modeling products. For example, you could obtain the sea surface temperature from the NOAA 4km AVHRR Pathfinder dataset. Then, assuming that a relatively standard multivariate statistical model was what you needed, MGET could help you fit a GLM, GAM, tree model, or random forest to model your response variable from the environmental values. Finally, if you wanted to produce a map predicting that response variable, MGET can predict a raster representing the response from a set of rasters representing the environment (e.g. SST images).

I hope that helps,

Jason