Regarding config["res"]: In their tutorials (tutorial1 and tutorial2), they either over-clustered and then merged clusters to achieve the target number or searched for a resolution initially. So, I added the resolution parameter to the config.
While testing, even if the initial resolution matches the expected clusters, we might still end up with more clusters. Hence, I included merging steps for both scenarios.
As all SOTIP spatial domain detection pipelines don't use any feature selection function, I didn't add n_genes.
Regarding
config["res"]: In their tutorials (tutorial1 and tutorial2), they either over-clustered and then merged clusters to achieve the target number or searched for a resolution initially. So, I added the resolution parameter to the config. While testing, even if the initial resolution matches the expected clusters, we might still end up with more clusters. Hence, I included merging steps for both scenarios.As all SOTIP spatial domain detection pipelines don't use any feature selection function, I didn't add
n_genes.