Fixed the error of get_results and get_results_rfc in cNMF modu…

…le. (#143) (#139)

omicverse/externel/STT/pl/_plot_tensor.py

@@ -7,7 +7,8 @@
 def plot_tensor_single(adata, adata_aggr = None, state = 'joint', 
                        attractor = None, basis = 'umap', color ='attractor', 
                        color_map = None, size = 20, alpha = 0.5, ax = None, 
-                       show = None, filter_cells = False, member_thresh = 0.05, density =2):
+                       show = None, filter_cells = False, member_thresh = 0.05, density =2,
+                       n_jobs = -1):
     """
     Function to plot a single tensor graph with assgined components
     
@@ -66,18 +67,22 @@ def plot_tensor_single(adata, adata_aggr = None, state = 'joint',
 
     if state == 'spliced':
         adata.layers['vs'] = velo[:,gene_select,1]
-        scv.tl.velocity_graph(adata, vkey = 'vs', xkey = 'Ms',n_jobs = -1)
+        scv.tl.velocity_graph(adata, vkey = 'vs', xkey = 'Ms',n_jobs = n_jobs)
         scv.pl.velocity_embedding_stream(adata, vkey = 'vs', basis=basis, color=color, title = title+','+'Spliced',color_map = color_map, size = size, alpha = alpha, ax = ax, show = show)
     if state == 'unspliced':
         adata.layers['vu'] = velo[:,gene_select,0]
-        scv.tl.velocity_graph(adata, vkey = 'vu', xkey = 'Mu',n_jobs = -1)
+        scv.tl.velocity_graph(adata, vkey = 'vu', xkey = 'Mu',n_jobs = n_jobs)
         scv.pl.velocity_embedding_stream(adata, vkey = 'vu',basis=basis, color=color, title = title+','+'Unspliced',color_map = color_map, size = size, alpha = alpha, ax = ax, show = show)
     if state == 'joint':
         print("check that the input includes aggregated object")
         #adata_aggr.layers['vj'] = np.concatenate((velo[:,gene_select,0],velo[:,gene_select,1]),axis = 1)
-        scv.tl.velocity_graph(adata_aggr, vkey = 'vj', xkey = 'Ms',n_jobs = -1)
+        scv.tl.velocity_graph(adata_aggr, vkey = 'vj', xkey = 'Ms',n_jobs = n_jobs)
         scv.pl.velocity_embedding_stream(adata_aggr, vkey = 'vj',basis=basis, color=color, title = title+','+'Joint',color_map = color_map, size = size, alpha = alpha, ax = ax, show = show, density =density)
 
+    del adata_copy
+    del adata_aggr_copy
+    import gc
+    gc.collect()
 
 
 def plot_tensor(adata, adata_aggr, list_state =['joint','spliced','unspliced'], list_attractor ='all', basis = 'umap',figsize = (8,8),hspace = 0.2,wspace = 0.2, color_map = None,size = 20,alpha = 0.5, filter_cells = False, member_thresh = 0.05, density =2):

omicverse/externel/STT/tl/_construct_landscape.py

@@ -32,7 +32,7 @@ def construct_landscape(sc_object,thresh_cal_cov = 0.3, scale_axis = 1.0, scale_
     mu_hat = sc_object.uns['da_out']['mu_hat']
     rho = sc_object.obsm['rho']
     projection = sc_object.obsm[coord_key][:,0:2]
-    p_hat=adata.uns['da_out']['P_hat']
+    p_hat=sc_object.uns['da_out']['P_hat']
 
 
     labels = np.argmax(rho,axis = 1)

omicverse/externel/cnmf/cnmf.py

@@ -1079,6 +1079,68 @@ def get_results_rfc(self,adata,result_dict,use_rep='STAGATE',cNMF_threshold=0.5)
         print("Single Tree:",score_c)
         print("Random Forest:",score_r)
 
+        adata.obs['cNMF_cluster_rfc']=[str(i) for i in rfc.predict(adata.obsm[use_rep])]
+        adata.obs['cNMF_cluster_clf']=[str(i) for i in clf.predict(adata.obsm[use_rep])]
+        print('cNMF_cluster_rfc is added to adata.obs')
+        print('cNMF_cluster_clf is added to adata.obs')
+
+    def get_results(self,adata,result_dict):
+        import pandas as pd
+        if result_dict['usage_norm'].columns[0] in adata.obs.columns:
+            #remove the columns if they already exist
+            #remove columns name starts with 'cNMF'
+            adata.obs = adata.obs.loc[:,~adata.obs.columns.str.startswith('cNMF')]
+        adata.obs = pd.merge(left=adata.obs, right=result_dict['usage_norm'], 
+                             how='left', left_index=True, right_index=True)
+        adata.var = pd.merge(left=adata.var,right=result_dict['gep_scores'].loc[adata.var.index],
+                             how='left', left_index=True, right_index=True)
+        df=adata.obs[result_dict['usage_norm'].columns].copy()
+        max_topic = df.idxmax(axis=1)
+        # 将结果添加到DataFrame中
+        adata.obs['cNMF_cluster'] = max_topic
+        print('cNMF_cluster is added to adata.obs')
+        print('gene scores are added to adata.var')
+
+    def get_results_rfc(self,adata,result_dict,use_rep='STAGATE',cNMF_threshold=0.5):
+        import pandas as pd
+        if result_dict['usage_norm'].columns[0] in adata.obs.columns:
+            #remove the columns if they already exist
+            #remove columns name starts with 'cNMF'
+            adata.obs = adata.obs.loc[:,~adata.obs.columns.str.startswith('cNMF')]
+        adata.obs = pd.merge(left=adata.obs, right=result_dict['usage_norm'], 
+                             how='left', left_index=True, right_index=True)
+        adata.var = pd.merge(left=adata.var,right=result_dict['gep_scores'].loc[adata.var.index],
+                             how='left', left_index=True, right_index=True)
+
+        import numpy as np
+        import pandas as pd
+        import matplotlib.pyplot as plt
+        from sklearn.tree import DecisionTreeClassifier
+        from sklearn.ensemble import RandomForestClassifier
+        from sklearn.model_selection import train_test_split
+
+        new_array = []
+        class_array = []
+        for i in range(1, result_dict['usage_norm'].shape[1] + 1):
+            data = adata[adata.obs[f'cNMF_{i}'] > cNMF_threshold].obsm[use_rep].toarray()
+            new_array.append(data)
+            class_array.append(np.full(data.shape[0], i))
+
+        new_array = np.concatenate(new_array, axis=0)
+        class_array = np.concatenate(class_array)
+
+        Xtrain, Xtest, Ytrain, Ytest = train_test_split(new_array,class_array,test_size=0.3)
+        clf = DecisionTreeClassifier(random_state=0)
+        rfc = RandomForestClassifier(random_state=0)
+        clf = clf.fit(Xtrain,Ytrain)
+        rfc = rfc.fit(Xtrain,Ytrain)
+        #查看模型效果
+        score_c = clf.score(Xtest,Ytest)
+        score_r = rfc.score(Xtest,Ytest)
+        #打印最后结果
+        print("Single Tree:",score_c)
+        print("Random Forest:",score_r)
+
         adata.obs['cNMF_cluster_rfc']=[str(i) for i in rfc.predict(adata.obsm[use_rep])]
         adata.obs['cNMF_cluster_clf']=[str(i) for i in clf.predict(adata.obsm[use_rep])]
         print('cNMF_cluster_rfc is added to adata.obs')

omicverse/pl/_single.py

@@ -133,7 +133,7 @@ def embedding(
 def cellproportion(adata:AnnData,celltype_clusters:str,groupby:str,
                        groupby_li=None,figsize:tuple=(4,6),
                        ticks_fontsize:int=12,labels_fontsize:int=12,ax=None,
-                       legend:bool=False):
+                       legend:bool=False,legend_awargs=None):
     """
     Plot cell proportion of each cell type in each visual cluster.
 
@@ -192,7 +192,7 @@ def cellproportion(adata:AnnData,celltype_clusters:str,groupby:str,
             bottoms+=test1['value'].values
         n+=1
     if legend!=False:
-        plt.legend(bbox_to_anchor=(1.05, -0.05), loc=3, borderaxespad=0,fontsize=10)
+        plt.legend(bbox_to_anchor=(1.05, -0.05), loc=3, borderaxespad=0,fontsize=10,**legend_awargs)
 
     plt.grid(False)
 
@@ -955,7 +955,7 @@ def plot_boxplots(  # pragma: no cover
 def cellstackarea(adata,celltype_clusters:str,groupby:str,
                        groupby_li=None,figsize:tuple=(4,6),
                        ticks_fontsize:int=12,labels_fontsize:int=12,ax=None,
-                       legend:bool=False):
+                       legend:bool=False,legend_awargs=None):
     """
     Plot the cell type percentage in each groupby category
     
@@ -1006,7 +1006,8 @@ def cellstackarea(adata,celltype_clusters:str,groupby:str,
         bottom += pivot_df[cell_type]
 
     if legend!=False:
-        plt.legend(bbox_to_anchor=(1.05, -0.05), loc=3, borderaxespad=0,fontsize=10)
+        plt.legend(bbox_to_anchor=(1.05, -0.05), loc=3, borderaxespad=0,
+                   fontsize=labels_fontsize,**legend_awargs)
 
     plt.grid(False)
 

omicverse_guide/docs/Release_notes.md

@@ -453,3 +453,18 @@ Support Raw Windows platform
 - Fixed type error of `float128` #134
 
 
+## v 1.6.7
+
+### Space Module
+
+- Added `n_jobs` argument to adjust thread in `extenel.STT.pl.plot_tensor_single`
+- Fixed an error in `extenel.STT.tl.construct_landscape`
+
+
+### Pl Module
+
+- Added `legend_awargs` to adjust the legend set in `pl.cellstackarea` and `pl.cellproportion`
+
+### Single Module
+
+- Fixed the error of `get_results` and `get_results_rfc` in `cNMF` module. (#143) (#139)