I could run through the PBMC tutorial data set with no problems, but got the ValueError when running desc.train on my own data. What did I do wrong?
ValueError: Cannot assign value to variable ' encoder_0/kernel:0': Shape mismatch.The variable shape (1141, 32), and the assigned value shape (2013, 32) are incompatible.
desc.version
'2.1.1'
tf.version
'2.12.0'
scanpy.version
'1.9.3'
My file after running
desc.normalize_per_cell(adata, counts_per_cell_after=1e4)
desc.log1p(adata)
sc.pp.highly_variable_genes(adata, min_mean=0.0125, max_mean=3, min_disp=0.5, subset=True)
desc.scale(adata, zero_center=True, max_value=3)
adata
AnnData object with n_obs × n_vars = 5141 × 1141
obs: 'BATCH', 'n_genes', 'n_counts'
var: 'n_cells', 'highly_variable', 'means', 'dispersions', 'dispersions_norm', 'mean', 'std'
uns: 'log1p', 'hvg'