Hello, sorry if the question is dumb... I wanted to use scib to evaluate some integration methods with my dataset, I had very difficult time figuring out the snakefile pipeline (I got the integration methods installed in the conda env but I just couldn't seem to figure how how to properly input my adata...) and still couldn't make it work so I resorted back to doing it one my one and I'll combine all the metrics when I'm done. However, I seem to have problem with even that...
>>>adata1 = scib.ig.scanorama(adata, batch="Dataset")
#adata is unitegrated, normalized raw counts
>>> adata1
AnnData object with n_obs × n_vars = 40714 × 4586
obs: 'orig.ident', 'Dataset', 'broad_celltype'
var: 'Column1'
obsm: 'X_scanorama', 'X_emb'
>>>scib.metrics.metrics(adata, adata1,batch_key='Dataset',label_key='broad_celltype')
0
NMI_cluster/label NaN
ARI_cluster/label NaN
ASW_label NaN
ASW_label/batch NaN
PCR_batch NaN
cell_cycle_conservation NaN
isolated_label_F1 NaN
isolated_label_silhouette NaN
graph_conn NaN
kBET NaN
iLISI NaN
cLISI NaN
hvg_overlap NaN
trajectory NaN
Sorry I know it's probably quite naive but I honestly don't know much about programming and spent quite some time reading and searching these question / issue threads and manual but still couldn't figure out why...
Hello, sorry if the question is dumb... I wanted to use scib to evaluate some integration methods with my dataset, I had very difficult time figuring out the snakefile pipeline (I got the integration methods installed in the conda env but I just couldn't seem to figure how how to properly input my adata...) and still couldn't make it work so I resorted back to doing it one my one and I'll combine all the metrics when I'm done. However, I seem to have problem with even that...
Sorry I know it's probably quite naive but I honestly don't know much about programming and spent quite some time reading and searching these question / issue threads and manual but still couldn't figure out why...