Add function to calculate batch ARI

[cbethell]

Is your feature request related to a problem? Please describe.
Per AlexsLemonade/scpca-downstream-analyses#352 (comment):
We are finding that we may have more than one use case for the integration performance metrics used in the sc-data-integration repo. That said, we should get these functions in a single place so that they can be utilized across multiple repos post-integration.

Describe the solution you'd like
Add a function for calculating batch ARI similar to the script in the sc-data-integration repo: https://github.com/AlexsLemonade/sc-data-integration/blob/main/scripts/utils/calculate-ARI.R

[allyhawkins]

This function should calculate the ARI between labels (either batch or cell type) and clustering results for a merged and integrated SCE object. PCs will be downsampled before clustering and calculating the ARI. This should be done a set number of times (default: 20) based on the number of reps specified.

As mentioned in #177, the setup to calculate the batch or cell type ARI is the same as setting up for calculating the ASW. So I think we should create a single wrapper function that grabs all the PCs and performs downsampling (see #177 (comment) for a longer description). If the metric specified when running that wrapper function is ARI, then we would run the ARI function that I'm describing below:

• Input:
  • PCs to cluster prior to computing ARI
  • vector of labels to compare to clustering results (e.g., cell type, batch, or for within-batch ARI, clusters from individual SCE object)
• Function:
  • Cluster PCs using graph-based clustering
  • calculate ARI between provided labels and clustering results
• Output:
  • A single ARI value

This function could also be used inside the function for calculating within-batch ARI mentioned in #176.
Note that the setup described here changes how we used to calculate batch ARI slightly. Instead of using k-means clustering and calculating the ARI across a range of k, this function would use graph-based clustering and output one ARI.

I'm going to tag folks that are out on Monday for their opinions. In the meantime @cbethell let me know if you have any additional thoughts.

[allyhawkins]

@sjspielman and @jashapiro could you please leave any additional thoughts or comments by EOD on Wednesday 6/28? Thank you!

[cbethell]

Addressed by #176