Radiance | master Radiance cvs source repo , used for the creation

The obvious reason would be that the difference between the two rasters is the same for all cells. With multiply and divide you would not notice that.

You can load your data "manually", using str.split and then feed the data to DataFrame. For example:

I have taken some time to parse your code.

In the large function, you use s3fs to interact with your cloud storage, and this works well with xarray.

However, in your main code, you use boto3 to list and open S3 files. These files retain a reference to the client object, which maintains a connection pool. That is the thing that cannot be pickled.

s3fs is designed to work with Dask, and ensures the picklebility of the filesystem instances and OpenFile objects. Since you already use it in one part, I would recommend using s3fs throughout (but I am, of course biased, since I am the main author).

Alternatively, you could pass just the file names (as strings), and not open anything until within the worker function. This would be "best practice" - you should load data in worker tasks, rather than loading in the client and passing the data.

Use combine='nested' instead.

From the Xarray documentation on combining by coords:

Attempt to auto-magically combine the given datasets into one by using dimension coordinates.

't' is not a dimension coordinate, so the xarray magic doesn't work in this case, because xarray's combine_by_coords looks for matching dimension coordinates between the imported netcdfs.

In this case you need to be more specific: use combine = 'nested' and specify the new dimension name with concat_dim='t'. As there is already a coordinate named 't' xarray will automatically promote it to dimension coordinate.

ds_sst = xr.open_mfdataset("OR_ABI-L1b-RadF*nc", concat_dim='t', combine='nested')

The resulting dataset looks like this.

With levels= you can tell contourf for which values you want a new level. In this, case choosing levels=0,1,2,...,9 seems to achieve what you want.

Note that the colors are decided in the call to contourf. The function colorbar just tries to visualize the colors that were used.

The code below first creates some random data between 0 and 20, with an outlier valued 50. The default call to contourf is compared to one were levels is set explicitly.

You need to use different texture units for your two textures. Even if each texture unit in the GL has a different binding slot for the different texture types (1D,2D,3D,Cube,1D Array, 2D Array, Cube Array and what ever), only one texture can be accessed via each texture unit during the draw call (meaning all texture accesses you're doing in all of the shader stages).

So it is fine if you have both a 2D texture and a cube map bount to the unit 0, and do one draw call which acceses unit 0 via sampler2D, and then switch to a program which accesses unit 0 via samplerCube, but it will fail as soon as a program tries to access unit 0 by more than one sampler type.

ImageGrid is attempting to maintain a square shape for pixels. If you have many more columns than lines, that may not be what you are trying to accomplish. Try passing aspect=False when constructing the ImageGrid.

Compare:

Text in MS word has extra information associated with it that can also be copied over when doing a copy paste into an html file or a CMS. Best practice for a copy and paste would be to try an intermediate text editor (like TextEdit for Mac or Note for Windows) or to convert the document to plain text before copying and pasting in your code.

You need to use nested repeater to display Benefits within the repeater as follows. See: Nested repeater.