The Köppen–Geiger climate classification made simpler (I hope so)

[Coggleton]

So, I went through the guide and did the initial precipitation map for January. However I'm unsure about some of the instructions with the transition map, as well as how far I should have certain zones go inland, so I haven't done any orographic lift effects or added in the penalty areas yet; would someone kindly be able to let me know if I'm roughly in the right direction? Much appreciated.

January Precipitation (Same color indicators as in guide/Charerg's legend):


January Pressure and Wind:

[Coggleton]

My apologies for doubleposting, but I think I might've run into an error with the new script; whenever I try to run it, it gives me the following error message:

"Plug-in 'Climate Generator' left image undo in inconsistent state, closing open undo groups."

Running it in the python console only gives the following error:
Traceback (most recent call last):
File "<input>", line 1, in <module>
RuntimeError: execution error

Is there any way to figure out what line of code the script crashed on?

I've tried reinstalling GIMP and installing Python, but to no avail. I have, however, been able to successfully run an older version of the script that's on post 334 of the thread. It can be seen as follows, with an added legend:



Going off of intuition, it seems odd that the tundra would go so far south on the coast right continent's top "finger". I'm also guessing that any internal tundra zones are due to mountains.

[Charerg]

The forum seems to have eaten the attachment (Attachment 125632) from your previous post, as it is not visible. Just to clarify, which version of GIMP are you running, and which version of the script you have an error with, the one posted in post #555?

[Coggleton]

I replaced the attachment with a text description of the error.

My GIMP version is 2.10.22, and it's the script in post #555 that I'm having issues with.

[Charerg]

I replaced the attachment with a text description of the error.

My GIMP version is 2.10.22, and it's the script in post #555 that I'm having issues with.

Can you post your source maps, so I can try to reproduce the error?

[Coggleton]

Of course; thanks in advance for taking a look at this.

JanTemp:


JulTemp:


JanPrec:


JulPrec:

[Charerg]

Well, the good news is that I had no problems with reproducing the error. The bad news is that it might take a while to figure this one out. Anyway, thanks for reporting the bug, I'll see if I can squash this one.

[Coggleton]

How does the script handle transparent pixels? While I thought I already filled them in, apparently I missed a couple. Then again, it still didn't fix the problem. New maps are pasted below.

JanTemp:


JulTemp:


JanPrec:


JulPrec:

[Charerg]

The current implementation doesn't tolerate any "off-colour" pixels at all (transparent or otherwise), which was the cause of the error. Though the matter is a bit more complicated, because apparently the script itself could paint the ocean in the wrong colour if you happened to have your opacity set at something other than 100% when running the script.

Anyway, I think I managed to figure this out now. I added a new line to the script that sets the opacity at 100% before painting the ocean and also new checks have been added so the script now tolerates off-colour pixels (it now assumes temperature to be 0° C and precipitation 1,0 mm if the colour is off). I'll post the updated version soon, once I've tested it a bit more with your source maps to make sure I didn't miss anything.

Edit:
Though I'd suggest making the maps a bit smaller than 22500x11250 for the new script, that could take a long while to process .

Edit2:
Ok, the script has now been updated (see post #555 for the updated version).

[Coggleton]

Heh, you have a point about the image size- unfortunately, scaling it down from the original size after i had made my elevation maps is what caused things to become janky (several pixels became various shades which I found unhelpful).

But the good news is, your code just ran and successfully executed.

EDIT: And after some color swapping, I'm proud to present the results below:


Besides the large amount of tundra up north, there's also considerably more desert than I expected. But in any case, I'm really grateful you threw the code together for this.