AttributeError: can't set attribute 'categories'

[BillyChen123]

I have not met this issue before. Why can this method have attribute error? Can you help me to fit it?

datasets = sc.read_h5ad(filepath)
datasets.obs['Batch'] = pd.CategoricalIndex(datasets.obs['Batch'])
datasets.obs['Batch2'] = pd.CategoricalIndex(datasets.obs['Batch2'])
datasets.obs['Group'] = pd.CategoricalIndex(datasets.obs['Group'])
sc.pp.normalize_total(datasets,target_sum= 1e4)
sc.pp.log1p(datasets)
sc.pp.highly_variable_genes(datasets)
#sc.pl.highly_variable_genes(datasets)
adata=desc.train(datasets,
        dims=[datasets.shape[1],64,32],
        tol=0.005,
        n_neighbors=10,
        batch_size=256,
        louvain_resolution=[0.8,1.0],# not necessarily a list, you can only set one value, like, louvain_resolution=1.0
        # save_dir=str(save_dir),
        do_tsne=True,
        learning_rate=200, # the parameter of tsne
        use_GPU=False,
        num_Cores=1, #for reproducible, only use 1 cpu
        num_Cores_tsne=4,
        save_encoder_weights=False,
        save_encoder_step=3,# save_encoder_weights is False, this parameter is not used
        use_ae_weights=False,
        do_umap=False) #if do_uamp is False, it will don't compute umap coordiate```

Here is the report!
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb 单元格 2 line 1
      [8](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=7) sc.pp.highly_variable_genes(datasets)
      [9](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=8) #sc.pl.highly_variable_genes(datasets)
---> [10](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=9) adata=desc.train(datasets,
     [11](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=10)         dims=[datasets.shape[1],64,32],
     [12](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=11)         tol=0.005,
     [13](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=12)         n_neighbors=10,
     [14](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=13)         batch_size=256,
     [15](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=14)         louvain_resolution=[0.8,1.0],# not necessarily a list, you can only set one value, like, louvain_resolution=1.0
     [16](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=15)         # save_dir=str(save_dir),
     [17](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=16)         do_tsne=True,
     [18](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=17)         learning_rate=200, # the parameter of tsne
     [19](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=18)         use_GPU=False,
     [20](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=19)         num_Cores=1, #for reproducible, only use 1 cpu
     [21](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=20)         num_Cores_tsne=4,
     [22](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=21)         save_encoder_weights=False,
     [23](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=22)         save_encoder_step=3,# save_encoder_weights is False, this parameter is not used
     [24](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=23)         use_ae_weights=False,
     [25](vscode-notebook-cell://ssh-remote%2B10.181.6.84/home/chenyz/Matrix_Factorization/rank/Untitled-1.ipynb#W5sdnNjb2RlLXJlbW90ZQ%3D%3D?line=24)         do_umap=False)

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/desc.py:303](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/desc.py:303), in train(data, dims, alpha, tol, init, louvain_resolution, n_neighbors, pretrain_epochs, batch_size, activation, actincenter, drop_rate_SAE, is_stacked, use_earlyStop, use_ae_weights, save_encoder_weights, save_encoder_step, save_dir, max_iter, epochs_fit, num_Cores, num_Cores_tsne, use_GPU, GPU_id, random_seed, verbose, do_tsne, learning_rate, perplexity, do_umap, kernel_clustering)
    301         print("Start to process resolution=",str(resolution))
    302         use_ae_weights=use_ae_weights if ith==0 else True
--> 303         res=train_single(data=data,
    304             dims=dims,
    305             alpha=alpha,
    306             tol=tol,
    307             init=init,
    308             louvain_resolution=resolution,
    309             n_neighbors=n_neighbors,
    310             pretrain_epochs=pretrain_epochs,
    311             epochs_fit=epochs_fit,
    312             batch_size=batch_size,
    313             activation=activation,
    314             actincenter=actincenter,
    315             drop_rate_SAE=drop_rate_SAE,
    316             is_stacked=is_stacked,
    317             use_earlyStop=use_earlyStop,
    318             use_ae_weights=use_ae_weights,
    319 	    save_encoder_weights=save_encoder_weights,
    320             save_encoder_step=save_encoder_step,
    321             save_dir=save_dir,
    322             max_iter=max_iter,
    323             num_Cores=num_Cores,
    324             num_Cores_tsne=num_Cores_tsne,
    325             use_GPU=use_GPU,
    326             GPU_id=GPU_id,
    327             random_seed=random_seed,
    328             verbose=verbose,
    329 	    do_tsne=do_tsne,
    330 	    learning_rate=learning_rate,
    331 	    perplexity=perplexity,
    332             do_umap=do_umap,
    333             kernel_clustering=kernel_clustering)
    334         #update adata
    335         data=res

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/desc.py:152](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/desc.py:152), in train_single(data, dims, alpha, tol, init, louvain_resolution, n_neighbors, pretrain_epochs, batch_size, activation, actincenter, drop_rate_SAE, is_stacked, use_earlyStop, use_ae_weights, save_encoder_weights, save_encoder_step, save_dir, max_iter, epochs_fit, num_Cores, num_Cores_tsne, use_GPU, GPU_id, random_seed, verbose, do_tsne, learning_rate, perplexity, do_umap, kernel_clustering)
    129 desc=DescModel(dims=dims,
    130           x=adata.X,
    131           alpha=alpha,
   (...)
    149           kernel_clustering=kernel_clustering
    150 )
    151 desc.compile(optimizer=SGD(0.01,0.9),loss='kld')
--> 152 Embeded_z,q_pred=desc.fit(maxiter=max_iter)
    153 print("The desc has been trained successfully!!!!!!")
    154 if verbose:

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/network.py:371](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/network.py:371), in DescModel.fit(self, maxiter)
    369 def fit(self,maxiter=1e4):
    370     if isinstance(self.epochs_fit,int):
--> 371         embedded_z,q=self.fit_on_all(maxiter=maxiter,epochs_fit=self.epochs_fit,save_encoder_step=self.save_encoder_step)
    372     else:
    373         import math

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/network.py:362](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/desc/models/network.py:362), in DescModel.fit_on_all(self, maxiter, epochs_fit, save_encoder_step)
    357 #load model
    358 #encoder=load_model("encoder.h5")
    359 #
    361 y0=pd.Series(y_pred,dtype='category')
--> 362 y0.cat.categories=range(0,len(y0.cat.categories))
    363 print("The final prediction cluster is:")
    364 x=y0.value_counts()

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/base.py:178](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/base.py:178), in NoNewAttributesMixin.__setattr__(self, key, value)
    172 if getattr(self, "__frozen", False) and not (
    173     key == "_cache"
    174     or key in type(self).__dict__
    175     or getattr(self, key, None) is not None
    176 ):
    177     raise AttributeError(f"You cannot add any new attribute '{key}'")
--> 178 object.__setattr__(self, key, value)

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/accessor.py:99](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/accessor.py:99), in PandasDelegate._add_delegate_accessors.<locals>._create_delegator_property.<locals>._setter(self, new_values)
     98 def _setter(self, new_values):
---> 99     return self._delegate_property_set(name, new_values)

File [~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/arrays/categorical.py:2867](https://vscode-remote+ssh-002dremote-002b10-002e181-002e6-002e84.vscode-resource.vscode-cdn.net/home/chenyz/Matrix_Factorization/rank/~/anaconda3/envs/MOBA_Billy/lib/python3.10/site-packages/pandas/core/arrays/categorical.py:2867), in CategoricalAccessor._delegate_property_set(self, name, new_values)
   2866 def _delegate_property_set(self, name: str, new_values):  # type: ignore[override]
-> 2867     return setattr(self._parent, name, new_values)

AttributeError: can't set attribute 'categories'

[goepp]

Hi,

I had the same issue while running the desc tutorial, with pandas=2.1.1 and desc=2.1.1.
I found a workaround inspired by theislab/scvelo#811 (comment).

1. In desc/models/network.py, I replaced line 318:
   y0.cat.categories=range(0,len(y0.cat.categories)) with

temp=y0.cat.categories.astype(str)
y0=y0.cat.set_categories(temp, rename=True)

2. In desc/models/desc.py I replaced line 159 :
   y_pred.cat.categories=list(range(len(y_pred.unique()))) with

temp=y_pred.cat.categories.astype(str)
y_pred=y_pred.cat.set_categories(temp, rename=True)

Then desc.train worked without problem.