Add Dask documentation#529
Conversation
SchneiderCode
left a comment
SchneiderCode
left a comment
There was a problem hiding this comment.
This is a great overview of utilizing Dask. Most of my comments below focus either on accessibility or personal preferences around readability/organization.
|
|
||
|  | ||
|
|
||
| ### Setting up a Local Cluster |
There was a problem hiding this comment.
We might want to add a Pre-Step that shows or references the Jupyter Notebooks configuration in Open OnDemand. This is important because the "n_workers=4" is likely related to the provided configurations (i.e. 4 cores, 4 hours).
This doesn't have to be a thorough discussion - we can also just have a link that points to the relevant OnDemand page in the RTD docs.
| - Uses CPU cores and memory | ||
| - Can run on the same node or across multiple nodes | ||
|
|
||
|  |
There was a problem hiding this comment.
As a general note - we'll need to add alternative text for all of the screenshots. Also, for screenshots that include notes / commands, we'll need to ensure that information is still accessible in the tutorial. For example, we could add a caption that details this information (if not already covered in the nearby instructions).
There was a problem hiding this comment.
If time allows, consider reformatting as a split column layout. With the description of Dask on the left and the image on the right. This could be done with sphinx grids:
https://sphinx-design.readthedocs.io/en/stable/grids.html
| ``` | ||
| This will generate a random array, and it will automatically create the tasks, and from there the sums will be parallelised. This is similar to what you would see in MPI, but much easier to implement. | ||
|
|
||
|  |
There was a problem hiding this comment.
Accessibility: Alt Text
Check if Dask provides alt-text for the visualization.
Co-authored-by: Michael Schneider <m.schneider.programmer@gmail.com>
Co-authored-by: Michael Schneider <m.schneider.programmer@gmail.com>
…s, and formatting improvements
2 participants
This PR adds new documentation for using Dask on Alpine, motivated by Issue 433. The goal of this addition is to provide a clear, end-to-end guide covering: