SURE-tools 2.1.11__py3-none-any.whl → 2.1.13__py3-none-any.whl

SURE/flow/plot_quiver.py

@@ -4,6 +4,7 @@ from sklearn.decomposition import PCA
 import scanpy as sc
 from matplotlib.colors import ListedColormap
 
+
 def plot_quiver(z_points, delta_z, method='pca', figsize=(6,4), dpi=200, 
                 subsample=None, color_by=None, colormap='viridis', 
                 arrow_scale=1.0, arrow_width=0.005, alpha=0.8):
@@ -23,96 +24,131 @@ def plot_quiver(z_points, delta_z, method='pca', figsize=(6,4), dpi=200,
         arrow_width: 箭头宽度
         alpha: 透明度
     """
-    # 数据采样
+    # 保存原始数据用于散点图
+    original_z_points = z_points.copy()
+    original_delta_z = delta_z.copy()
+    original_color_by = color_by.copy() if color_by is not None else None
+    
+    # 数据采样（仅用于quiver）
+    quiver_indices = None
     if subsample is not None and len(z_points) > subsample:
-        idx = np.random.choice(len(z_points), subsample, replace=False)
-        z_points = z_points[idx]
-        delta_z = delta_z[idx]
+        quiver_indices = np.random.choice(len(z_points), subsample, replace=False)
+        z_points_quiver = z_points[quiver_indices]
+        delta_z_quiver = delta_z[quiver_indices]
         if color_by is not None:
-            color_by = color_by[idx]
+            color_by_quiver = color_by[quiver_indices]
+        else:
+            color_by_quiver = None
+        color_by_quiver=None
+    else:
+        z_points_quiver = z_points
+        delta_z_quiver = delta_z
+        color_by_quiver = color_by
+        color_by_quiver=None
     
-    # 降维投影
+    # 降维投影（使用所有数据）
     if method == 'variance':
-        variances = np.var(z_points, axis=0)
+        variances = np.var(original_z_points, axis=0)
         dims = np.argsort(variances)[-2:]
-        z_2d = z_points[:, dims]
-        delta_z_2d = delta_z[:, dims]
+        z_2d_all = original_z_points[:, dims]
+        z_2d_quiver = z_points_quiver[:, dims]
+        delta_z_2d = delta_z_quiver[:, dims]
         dim_names = [f'z[{d}]' for d in dims]
         
     elif method == 'pca':
         pca = PCA(n_components=2)
-        z_2d = pca.fit_transform(z_points)
-        delta_z_2d = pca.transform(z_points + delta_z) - z_2d
+        z_2d_all = pca.fit_transform(original_z_points)
+        z_2d_quiver = pca.transform(z_points_quiver)
+        delta_z_2d = pca.transform(z_points_quiver + delta_z_quiver) - z_2d_quiver
         dim_names = ['PC1', 'PC2']
         
     elif method == 'manual':
         dims = [0, 1]
-        z_2d = z_points[:, dims]
-        delta_z_2d = delta_z[:, dims]
+        z_2d_all = original_z_points[:, dims]
+        z_2d_quiver = z_points_quiver[:, dims]
+        delta_z_2d = delta_z_quiver[:, dims]
         dim_names = [f'z[{d}]' for d in dims]
         
     elif method == 'umap':
-        ad = sc.AnnData(np.vstack([z_points, z_points+delta_z]))
+        # 使用所有数据进行UMAP降维
+        ad = sc.AnnData(np.vstack([original_z_points, original_z_points+original_delta_z]))
         sc.pp.neighbors(ad)
         sc.tl.umap(ad)
-        z_2d = ad[:z_points.shape[0]].obsm['X_umap']
-        delta_z_2d = ad[z_points.shape[0]:].obsm['X_umap'] - z_2d
+        z_2d_all = ad[:len(original_z_points)].obsm['X_umap']
+        z_2d_quiver = z_2d_all[quiver_indices] if quiver_indices is not None else z_2d_all
+        delta_z_2d = ad[len(original_z_points):].obsm['X_umap'][quiver_indices] - z_2d_quiver if quiver_indices is not None else ad[len(original_z_points):].obsm['X_umap'] - z_2d_all
         dim_names = ['UMAP1', 'UMAP2']
     
-    # 颜色处理
-    if color_by is not None:
+    # 颜色处理（散点图使用所有数据，quiver使用采样数据）
+    if original_color_by is not None:
         if isinstance(colormap, str):
             cmap = plt.get_cmap(colormap)
         else:
             cmap = colormap
         
-        if color_by.dtype.kind in ['i', 'f']:  # 数值型标签
-            colors = cmap(color_by / max(color_by.max(), 1e-8))
+        if original_color_by.dtype.kind in ['i', 'f']:  # 数值型标签
+            # 所有数据的颜色
+            colors_all = cmap(original_color_by / max(original_color_by.max(), 1e-8))
+            # quiver数据的颜色
+            if color_by_quiver is not None:
+                colors_quiver = cmap(color_by_quiver / max(color_by_quiver.max(), 1e-8))
+            else:
+                colors_quiver = 'blue'
             cbar_label = 'Numeric Label'
         else:  # 类别型标签
-            unique_labels = np.unique(color_by)
+            unique_labels = np.unique(original_color_by)
             color_map = {label: cmap(i/len(unique_labels)) 
                         for i, label in enumerate(unique_labels)}
-            colors = [color_map[label] for label in color_by]
+            colors_all = [color_map[label] for label in original_color_by]
+            if color_by_quiver is not None:
+                colors_quiver = [color_map[label] for label in color_by_quiver]
+            else:
+                colors_quiver = 'blue'
             cbar_label = 'Class Label'
     else:
-        colors = 'blue'
+        colors_all = 'gray'
+        colors_quiver = 'blue'
     
     # 绘制
     plt.figure(figsize=figsize, dpi=dpi)
     
-    # 绘制quiver（分颜色组绘制以获得正确图例）
-    if color_by is not None and isinstance(color_by[0], str):
-        for label in np.unique(color_by):
-            mask = color_by == label
-            plt.quiver(z_2d[mask, 0], z_2d[mask, 1],
+    # 1. 首先绘制所有数据点的散点图
+    plt.scatter(z_2d_all[:, 0], z_2d_all[:, 1], 
+               c=colors_all, alpha=0.3, s=5, label='All data points')
+    
+    # 2. 绘制quiver（仅采样数据）
+    if color_by_quiver is not None and isinstance(color_by_quiver[0], str):
+        for label in np.unique(color_by_quiver):
+            mask = color_by_quiver == label
+            plt.quiver(z_2d_quiver[mask, 0], z_2d_quiver[mask, 1],
                       delta_z_2d[mask, 0], delta_z_2d[mask, 1],
                       angles='xy', scale_units='xy', scale=1.0/arrow_scale,
-                      color=colors[mask], width=arrow_width, alpha=alpha,
-                      label=str(label))
+                      color=colors_quiver[mask], width=arrow_width, alpha=alpha,
+                      label=f'{label} (movement)')
         plt.legend()
     else:
-        q = plt.quiver(z_2d[:, 0], z_2d[:, 1],
+        q = plt.quiver(z_2d_quiver[:, 0], z_2d_quiver[:, 1],
                       delta_z_2d[:, 0], delta_z_2d[:, 1],
                       angles='xy', scale_units='xy', scale=1.0/arrow_scale,
-                      color=colors, width=arrow_width, alpha=alpha)
+                      color=colors_quiver, width=arrow_width, alpha=alpha,
+                      label='Movement vectors')
         
         # 添加颜色条（数值型标签）
-        if color_by is not None and color_by.dtype.kind in ['i', 'f']:
+        if original_color_by is not None and original_color_by.dtype.kind in ['i', 'f']:
             plt.colorbar(plt.cm.ScalarMappable(
-                norm=plt.Normalize(color_by.min(), color_by.max()),
+                norm=plt.Normalize(original_color_by.min(), original_color_by.max()),
                 cmap=cmap), label=cbar_label)
     
     # 美化图形
-    plt.scatter(z_2d[:, 0], z_2d[:, 1], c='gray', alpha=0.3, s=5)
     plt.xlabel(dim_names[0])
     plt.ylabel(dim_names[1])
-    plt.title(f"Latent Space Movement ({method} projection)")
+    plt.title(f"Latent Space Movement ({method} projection)\n{len(z_points_quiver)}/{len(original_z_points)} points with vectors")
     plt.grid(alpha=0.2)
+    plt.legend()
     plt.tight_layout()
     plt.show()
     
-    return z_2d, delta_z_2d
+    return z_2d_all, z_2d_quiver, delta_z_2d
 
 def plot_quiver_old(z_points, delta_z, method='pca', figsize=(6,4), dpi=200):
     """

{sure_tools-2.1.11.dist-info → sure_tools-2.1.13.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: SURE-tools
-Version: 2.1.11
+Version: 2.1.13
 Summary: Succinct Representation of Single Cells
 Home-page: https://github.com/ZengFLab/SURE
 Author: Feng Zeng

{sure_tools-2.1.11.dist-info → sure_tools-2.1.13.dist-info}/RECORD

@@ -11,7 +11,7 @@ SURE/codebook/__init__.py,sha256=2T5gjp8JIaBayrXAnOJYSebQHsWprOs87difpR1OPNw,243
 SURE/codebook/codebook.py,sha256=ZlN6gRX9Gj2D2u3P5KeOsbZri0MoMAiJo9lNeL-MK-I,17117
 SURE/flow/__init__.py,sha256=rsAjYsh1xVIrxBCuwOE0Q_6N5th1wBgjJceV0ABPG3c,183
 SURE/flow/flow_stats.py,sha256=3F3waCuEbIQ7bsiGga4cUvJphYdWA307SyGwEh8EzM8,10514
-SURE/flow/plot_quiver.py,sha256=FyFc4lSRuY2bMmPMQtPsfwAJ7xhw7XD_e_OCVWa4ZeM,6083
+SURE/flow/plot_quiver.py,sha256=wC42yLKfXYOGrrd4u9AbZS_6QiPUJ9QBXaOso55LjdA,8110
 SURE/flow/quiver.py,sha256=_euFqSaRrDoZ_oOabOx20LOoUTJ__XPhLW-vzLNQfAo,1859
 SURE/perturb/__init__.py,sha256=ouxShhbxZM4r5Gf7GmKiutrsmtyq7QL8rHjhgF0BU08,32
 SURE/perturb/perturb.py,sha256=CqO3xPfNA3cG175tadDidKvGsTu_yKfJRRLn_93awKM,3303
@@ -19,9 +19,9 @@ SURE/utils/__init__.py,sha256=Htqv4KqVKcRiaaTBsR-6yZ4LSlbhbzutjNKXGD9-uds,660
 SURE/utils/custom_mlp.py,sha256=07TYX1HgxfEjb_3i5MpiZfNhOhx3dKntuwGkrpteWiM,7036
 SURE/utils/queue.py,sha256=E_5PA5EWcBoGAZj8BkKQnkCK0p4C-4-xcTPqdIXaPXU,1892
 SURE/utils/utils.py,sha256=IUHjDDtYaAYllCWsZyIzqQwaLul6fJRvHRH4vIYcR-c,8462
-sure_tools-2.1.11.dist-info/licenses/LICENSE,sha256=TFHKwmrAViXQbSX5W-NDItkWFjm45HWOeUniDrqmnu0,1065
-sure_tools-2.1.11.dist-info/METADATA,sha256=aUC_xD4vGERur0cFsccW59T1ufKKN7mbSDMHMJXHBko,2651
-sure_tools-2.1.11.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-sure_tools-2.1.11.dist-info/entry_points.txt,sha256=-nJI8rVe_qqrR0HmfAODzj-JNfEqCcSsyVh6okSqyHk,83
-sure_tools-2.1.11.dist-info/top_level.txt,sha256=BtFTebdiJeqra4r6mm-uEtwVRFLZ_IjYsQ7OnalrOvY,5
-sure_tools-2.1.11.dist-info/RECORD,,
+sure_tools-2.1.13.dist-info/licenses/LICENSE,sha256=TFHKwmrAViXQbSX5W-NDItkWFjm45HWOeUniDrqmnu0,1065
+sure_tools-2.1.13.dist-info/METADATA,sha256=ZcsPEH_ed7pp9FaJCAJClFC4EeFY3l7TGgEFg7yRTYM,2651
+sure_tools-2.1.13.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+sure_tools-2.1.13.dist-info/entry_points.txt,sha256=-nJI8rVe_qqrR0HmfAODzj-JNfEqCcSsyVh6okSqyHk,83
+sure_tools-2.1.13.dist-info/top_level.txt,sha256=BtFTebdiJeqra4r6mm-uEtwVRFLZ_IjYsQ7OnalrOvY,5
+sure_tools-2.1.13.dist-info/RECORD,,