Add pipeline module and ase function #814
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
… only in it. One major aspect of this seemingly innocuous change is that we want to be able to seed the randomized svd. This in turn has had a cascade of changes in most of the spectral embeddings and their tests. It wasn't until I was fixing these tests that I found out we weren't writing unittests using the unittest module, but instead relying solely on pytest to find and execute all of our tests. It has done this well for the CLI, but it breaks a lot of IDE test runner code, making it harder to debug tests since none are actually being executed in intellij/pycharm/etc.
|
❌ Deploy Preview for graspologic failed. 🔨 Explore the source changes: af18c2d 🔍 Inspect the deploy log: https://app.netlify.com/sites/graspologic/deploys/61086c774735200008aac149 |
daxpryce
requested review from
bdpedigo,
bryantower,
jonmclean,
nicaurvi and
j1c
… instead to use nx.utils.graphs_equal, but that doens't exist in 2.5 (or earlier).
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
bdpedigo
reviewed
Jul 29, 2021
- Moved preconditions and module and associated test to top level - Adjusted documentation for more clarity and detail - Fixed a test in omni that doens't fail now, but will likely fail later after we fix the outstanding issue in in select_svd when "randomized" algorithm is used. Namely, a lot of unit tests if we fix things correctly, so instead we're defaulting to sklearn's randomized svd behavior, and will fix this properly in another issue.
…the preconditions module to the top level, but also some further details aroudn the embed submodule of the pipeline package
bryantower
approved these changes
Aug 3, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
So, this is the first commits to the pipeline module. It includes a few QOL items like an int-ifying graph builder class that will generate an
nx.Graphornx.DiGraph, a preconditions module for type checking (probably could end up in a utils module in the top level), an Embeddings return object that makes it trivial to correlate a node id to it's respective tensor, and an implementation of ASE-as-a-function.Of particular note is that it does not use the built-in elbow finder yet, instead behaving more similarly to the topologic approach of using a combination of fixed dimensionality requested and the elbow finder on that. We've discussed addressing our
graspologic.embed.AdjacencySpectralEmbedelbow-finding concerns in a meeting and we're reasonably confident we can have a consistent behavior in both in the future, but for now this works well enough.