Wentao Li

@rcomer Thank you for your continued effort on this issue; your test case is superbly done. The test case I added was specifically for testing the function `path_to_geos`, whereas yours...

Okay, I understand. Thank you very much!

Sorry, I've found that the modified code introduced a new bug in actual use, and I'm currently addressing it. I will provide a more detailed description later.

The situation we are now facing is more complex than it was at the beginning. I've discovered a new issue: if we use [this data](https://drive.google.com/file/d/11ke5iViG3OBLIxQ8WFUMAdTmAX6QKKN2/view?usp=sharing) to plot a graph, it...

@greglucas I have referred to the example in [contourpy/contourpy#377](https://github.com/contourpy/contourpy/issues/377), but unfortunately, based on my actual testing, I found that the `make_valid` function does not guarantee the return type as a...

I'm not sure if I'm misunderstanding the `contour/contourf` function, but I've always thought that each `Path` in the collection object returned by the matplotlib `contour/contourf` function represents a separate polygon,...

I understand now, so `path_to_geos` is a very general function that handles all Path objects built with matplotlib, converting them into shapely geometric objects for further projection transformations and other...

> The +/- buffering seems like a hack to me and may not be suitable for all resolutions / scales. One other thought that comes to mind is that this...

@greglucas Regarding your suggestions, I have conducted the following work: 1. To understand if it is possible to directly set `is_inside` to False when a `GEOSException` occurs, I checked the...

> If so I wonder if we could use the output of `buffer(0)` to define `is_inside`, but not actually replace the original polygon. I just tried locally to only judge...