evapotranspiration | Ruby library for calculating reference crop

According to the "Layers" section of the dataset docs, the class assignment in fill values 32761 through 32767 might be inaccurate. They are not included in the Earth Engine assets.

Users are asked to disregard the class assignments for Fill Values 32761 through 32767 irrespective of the assigned quality for these pixels. Class assignments may be incorrect and users are advised to exercise caution while using the product in their application. See known issue PM_MOD16_18074 for further information.

I would suggest using rollmean() from zoo and a tidyverse approach with across() from dplyr:

The last part of your example did not work and I changed it to this (for one year)

The problem with your first code snippet is that the text you're passing as a parameter to the HTTP call is too long, if you print the response object you'll see:

that corresponds to 414 URI Too Long Reference

If you pass a smaller text, dbpedia-spotlight will be able to annotate the entities for you.

For the second code that you put, you have two problems, the first one is that dbpedia-spotlight may respond with 403 status after consecutive calls to the annotate service, to check that I suggest you to do:

You were quite close to the solution. Your final data contains everything to get your plot.

Basically, you can use facet_grid to separate both "Year" and "Model" variables and geom_sf that will use the column "geometry" of your dataset to plot the shape (more informations here: https://ggplot2.tidyverse.org/reference/ggsf.html). Passing the value as a filling into the aes will make the rest.

Here, I used scale_fill_gradient to set the color and few functions to get the plot as close as your desired plot:

You need to keep a list of tables rather than appending all of the information into a single list. Then you can use zip() to take a single header and table at a time to write it to your output CSV file. This is a better approach than trying to use range().

You should use resampledoc with the frequency with one month. Then:

So your source data identified earlier is from a dict called grouped (think if you could put that back in and show an example that be great)

From group you want to get all the keys as headers and values as values but replacing all the symbols you do not need.

The code below does that for you starting from your grouped dict and stores your header and value into 2 seperate list: