sciv 0.0.100__tar.gz → 0.0.102__tar.gz

{sciv-0.0.100 → sciv-0.0.102}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sciv
-Version: 0.0.100
+Version: 0.0.102
 Summary: Unveiling the pivotal cell types involved in variant function regulation at a single-cell resolution
 Project-URL: github, https://github.com/YuZhengM/sciv
 Author-email: Zheng-Min Yu <yuzmbio@163.com>
@@ -48,6 +48,7 @@ Requires-Dist: numba==0.60.0
 Requires-Dist: numpy==1.26.4
 Requires-Dist: palettable==3.3.3
 Requires-Dist: pandas==1.5.3
+Requires-Dist: psutil==7.2.1
 Requires-Dist: pycomplexheatmap==1.6.1
 Requires-Dist: pynndescent==0.5.13
 Requires-Dist: scanpy==1.11.5

{sciv-0.0.100 → sciv-0.0.102}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["hatchling"]
 build-backend = "hatchling.build"
 [project]
 name = "sciv"
-version = "0.0.100"
+version = "0.0.102"
 authors = [
     { name = "Zheng-Min Yu", email = "yuzmbio@163.com" },
 ]
@@ -48,7 +48,8 @@ dependencies = [
     "louvain==0.8.2",
     "numba==0.60.0",
     "pynndescent==0.5.13",
-    "fa2-modified==0.4"
+    "fa2-modified==0.4",
+    "psutil==7.2.1"
 ]
 classifiers = [
     "Programming Language :: Python :: 3",

{sciv-0.0.100 → sciv-0.0.102}/requirements.txt

@@ -28,4 +28,5 @@ louvain==0.8.2
 snapatac2==2.7.1
 numba==0.60.0
 pynndescent==0.5.13
-fa2-modified==0.4
+fa2-modified==0.4
+psutil==7.2.1

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/model/_core_.py

@@ -21,7 +21,7 @@ __name__: str = "model_core"
 
 
 def _run_random_walk_(random_walk: RandomWalk, is_ablation: bool, is_simple: bool) -> AnnData:
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     if not random_walk.is_run_core:
         random_walk.run_core()
@@ -57,7 +57,7 @@ def _run_random_walk_(random_walk: RandomWalk, is_ablation: bool, is_simple: boo
         if not random_walk.is_run_en_ablation_m_knn:
             random_walk.run_en_ablation_m_knn()
 
-    random_walk.elapsed_time += time.time() - start_time
+    random_walk.elapsed_time += time.perf_counter() - start_time
 
     return random_walk.trs_adata
 
@@ -181,7 +181,7 @@ def core(
     """
 
     # start time
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     if len(variants.keys()) == 0:
         ul.log(__name__).error("The number of mutations is empty.")
@@ -591,7 +591,7 @@ def core(
     random_walk_time = random_walk.elapsed_time
 
     # end time
-    elapsed_time = time.time() - start_time
+    elapsed_time = time.perf_counter() - start_time
     step_time = poisson_vi_time + overlap_time + init_score_time + smknn_time + random_walk_time
 
     if elapsed_time < step_time:

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_bar_.py

@@ -76,6 +76,7 @@ def two_bar(
     bottom: float = 0,
     rotation: float = 65,
     text_left_move: float = 0.15,
+    y_limit: Tuple[float, float] = (0, 1),
     title: str = None,
     output: path = None,
     show: bool = True,
@@ -91,6 +92,9 @@ def two_bar(
 
     ax.legend()
 
+    ax.set_ylim(y_limit)
+
+    ax.set_xticks(range(len(ax_x)))
     ax.set_xticklabels(labels=list(ax_x), rotation=rotation)
 
     # Draw numerical values
@@ -103,6 +107,13 @@ def two_bar(
             color=text_color
         )
 
+    for spine in ["top", "left", "right", "bottom"]:
+        ax.spines[spine].set_linewidth(1)
+
+    ax.spines['bottom'].set_linewidth(1)
+    ax.grid(axis='y', ls='--', c='gray')
+    ax.set_axisbelow(True)
+
     plot_end(fig, title, x_name, y_name, output, show, close)
 
 
@@ -123,6 +134,7 @@ def class_bar(
     rotation: float = 65,
     title: str = None,
     text_left_move: float = 0.15,
+    y_limit: Tuple[float, float] = (0, 1),
     output: path = None,
     show: bool = True,
     close: bool = False,
@@ -159,6 +171,7 @@ def class_bar(
         bottom=bottom,
         rotation=rotation,
         text_left_move=text_left_move,
+        y_limit=y_limit,
         title=title,
         output=output,
         show=show,
@@ -185,6 +198,7 @@ def bar_trait(
     rotation: float = 65,
     title: str = None,
     text_left_move: float = 0.15,
+    y_limit: Tuple[float, float] = (0, 1),
     output: path = None,
     show: bool = True,
     close: bool = False,
@@ -216,6 +230,7 @@ def bar_trait(
             bottom=bottom,
             rotation=rotation,
             text_left_move=text_left_move,
+            y_limit=y_limit,
             text_color=text_color,
             output=os.path.join(output, f"cell_{trait_}_enrichment_bar.pdf") if output is not None else None,
             show=show,
@@ -224,6 +239,7 @@ def bar_trait(
         )
 
     trait_list = list(set(trait_df[trait_column_name]))
+
     # judge trait
     if trait_name != "All" and trait_name not in trait_list:
         ul.log(__name__).error(
@@ -237,8 +253,10 @@ def bar_trait(
 
     # plot
     if trait_name == "All":
+
         for trait in trait_list:
             trait_plot(trait_=trait, cell_df_=trait_df)
+
     else:
         trait_plot(trait_name, trait_df)
 
@@ -376,10 +394,9 @@ def bar_significance(
         y=y,
         hue=legend,
         hue_order=hue_order,
-        ci=ci,
+        errorbar=('ci', ci),
         capsize=capsize,
-        errwidth=line_width,
-        errcolor=errcolor,
+        err_kws={'color': errcolor, 'linewidth': line_width},
         ax=ax,
         palette=palette,
         edgecolor=errcolor,
@@ -437,7 +454,7 @@ def bar_significance(
     ax.set_axisbelow(True)
 
     if x_rotation != 0:
-        ax.set_xticklabels(labels=ax.get_xticklabels(), rotation=x_rotation)
+        ax.tick_params(axis='x', rotation=x_rotation)
 
     plt.legend(loc='upper left', bbox_to_anchor=(0.0, legend_gap), ncol=2)
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_box_.py

@@ -52,7 +52,6 @@ def box_base(
 
     if "color" in df_columns:
         new_df_color: DataFrame = df.groupby(group_columns, as_index=False)["color"].first()
-
         new_df = new_df.merge(new_df_color, how="left", on=clusters)
 
     colors: list = []
@@ -61,22 +60,31 @@ def box_base(
     if is_sort:
         new_df.sort_values([value], ascending=False, inplace=True)
         y_names: Union[list, None] = list(new_df[clusters])
+
         if "color" in df_columns:
-            colors = new_df["color"]
+            colors = list(new_df["color"])
+
     else:
         new_df.index = new_df[clusters]
+
         if order_names is not None:
             y_names: list = order_names
+
             if "color" in df_columns:
+
                 for i in order_names:
+
                     for j, c in zip(new_df[clusters], new_df["color"]):
+
                         if i == j:
                             colors.append(c)
                             break
+
         else:
             y_names = new_df[clusters]
+
             if "color" in df_columns:
-                colors = new_df["color"]
+                colors = list(new_df["color"])
 
     # scatter
     sns.boxplot(
@@ -103,6 +111,7 @@ def box_base(
         line.set_linewidth(line_width)
 
     # set coordinate
+    ax.set_xticks(range(len(y_names)))
     ax.set_xticklabels(labels=y_names, rotation=rotation)
     ax.spines['top'].set_linewidth(line_width)
     ax.spines['bottom'].set_linewidth(line_width)

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_core_.py

@@ -1,7 +1,7 @@
 # -*- coding: UTF-8 -*-
 
 import os
-from typing import Tuple, Literal, Optional, Union
+from typing import Tuple, Literal, Optional, Union, Any
 
 import pandas as pd
 from anndata import AnnData
@@ -314,9 +314,11 @@ def rate_bar_plot(
     rotation: float = 65,
     title: str = None,
     text_left_move: float = 0.15,
+    y_limit: Tuple[float, float] = (0, 1),
     plot_output: path = None,
     show: bool = True,
-    close: bool = False
+    close: bool = False,
+    **kwargs: Any
 ) -> None:
 
     if dir_name is not None:
@@ -349,9 +351,11 @@ def rate_bar_plot(
         bottom=bottom,
         rotation=rotation,
         text_left_move=text_left_move,
+        y_limit=y_limit,
         output=new_path if plot_output is not None else None,
         show=show,
-        close=close
+        close=close,
+        **kwargs
     )
 
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_graph_.py

@@ -143,8 +143,10 @@ def communities_graph(
     partition: list = [0 for _ in range(g.number_of_nodes())]
 
     for c_i, nodes in enumerate(communities):
+
         for i in nodes:
             partition[i] = type_colors[start_color_index + color_index * color_step_size + c_i]
+
         color_index += 1
 
     pos = nx.spring_layout(g)

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_kde_.py

@@ -42,7 +42,7 @@ def kde(
     # Random sampling
     if axis == -1:
         matrix = down_sampling_data(data.X, sample_number)
-        sns.kdeplot(matrix, shade=True, fill=True)
+        sns.kdeplot(matrix, fill=True)
     elif axis == 0:
         col_number = data.shape[1]
         if data.shape[0] * data.shape[1] > sample_number:
@@ -50,10 +50,10 @@ def kde(
 
             for i in tqdm(range(col_number)):
                 _vector_ = down_sampling_data(data.X[:, i], row_number)
-                sns.kdeplot(np.array(_vector_).flatten(), shade=True, fill=True, **kwargs)
+                sns.kdeplot(np.array(_vector_).flatten(), fill=True, **kwargs)
         else:
             for i in tqdm(range(col_number)):
-                sns.kdeplot(np.array(data.X[:, i]).flatten(), shade=True, fill=True, **kwargs)
+                sns.kdeplot(np.array(data.X[:, i]).flatten(), fill=True, **kwargs)
 
         if is_legend:
             ax.legend(list(adata.var.index))
@@ -65,10 +65,10 @@ def kde(
 
             for i in tqdm(range(row_number)):
                 _vector_ = down_sampling_data(data.X[i, :], col_number)
-                sns.kdeplot(np.array(_vector_).flatten(), shade=True, fill=True, **kwargs)
+                sns.kdeplot(np.array(_vector_).flatten(), fill=True, **kwargs)
         else:
             for i in tqdm(range(row_number)):
-                sns.kdeplot(np.array(data.X[i, :]).flatten(), shade=True, fill=True, **kwargs)
+                sns.kdeplot(np.array(data.X[i, :]).flatten(), fill=True, **kwargs)
 
         if is_legend:
             ax.legend(list(adata.obs.index))

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_line_.py

@@ -119,7 +119,7 @@ def stability_line(
         locator = mdates.DayLocator(interval=1)
         chart.xaxis.set_major_locator(locator)
 
-        ax.set_xticklabels(ax.get_xticklabels(), rotation=x_name_rotation)
+        ax.tick_params(axis='x', rotation=x_name_rotation)
     else:
         plt.xticks(x_ticks, rotation=x_name_rotation)
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_pie_.py

@@ -115,6 +115,7 @@ def pie_label(
         startangle=90,
         labeldistance=label_distance,
         pctdistance=pct_distance,
+        wedgeprops=dict(linewidth=0),
         **kwargs
     )
     ax.pie(
@@ -122,7 +123,7 @@ def pie_label(
         colors=['white'],
         radius=radius,
         startangle=90,
-        wedgeprops=dict(width=radius, edgecolor='w'),
+        wedgeprops=dict(width=radius, edgecolor='w', linewidth=0),
         **kwargs
     )
     ax.text(0, 0, "{:.2f}%".format(top_x[0] / top_sum * 100), ha='center', va='center', fontsize=fontsize)

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/plot/_violin_.py

@@ -48,7 +48,8 @@ def violin_base(
 
     if hue is not None and hue not in df_columns:
         ul.log(__name__).error(
-            f"The `hue` ({hue}) parameter must be in the `df` parameter data column name ({df_columns})")
+            f"The `hue` ({hue}) parameter must be in the `df` parameter data column name ({df_columns})"
+        )
         raise ValueError(f"The `hue` ({hue}) parameter must be in the `df` parameter data column name ({df_columns})")
 
     fig, ax = plot_start(width, height, bottom, output, show)
@@ -67,20 +68,29 @@ def violin_base(
     if is_sort:
         new_df.sort_values([value], ascending=False, inplace=True)
         y_names: Union[list, None] = list(new_df[clusters])
+
         if "color" in df_columns:
             colors = list(new_df["color"])
+
     else:
         new_df.index = new_df[clusters]
+
         if order_names is not None:
             y_names: list = order_names
+
             if "color" in df_columns:
+
                 for i in order_names:
+
                     for j, c in zip(new_df[clusters], new_df["color"]):
+
                         if i == j:
                             colors.append(c)
                             break
+
         else:
             y_names = list(new_df[clusters])
+
             if "color" in df_columns:
                 colors = list(new_df["color"])
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/preprocessing/_scanpy_.py

@@ -1,4 +1,5 @@
 # -*- coding: UTF-8 -*-
+
 import time
 import warnings
 from typing import Optional, Literal
@@ -47,7 +48,7 @@ def filter_data(
     """
 
     # start time
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     import scanpy as sc
 
@@ -109,7 +110,7 @@ def filter_data(
     )
     ul.log(__name__).info(f"Size of filtered scATAC-seq data: {filter_adata.shape}")
     filter_adata.uns["step"] = 0
-    filter_adata.uns["elapsed_time"] = time.time() - start_time
+    filter_adata.uns["elapsed_time"] = time.perf_counter() - start_time
 
     return filter_adata
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/preprocessing/_scvi_.py

@@ -50,7 +50,7 @@ def poisson_vi(
     """
     ul.log(__name__).info("Start PoissonVI")
 
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     import scvi
     import scanpy as sc
@@ -241,7 +241,7 @@ def poisson_vi(
     da_peaks_adata.layers["emp_prob1"] = matrix_ep1
     da_peaks_adata.uns["latent_name"] = latent_name
     da_peaks_adata.uns["dp_delta"] = dp_delta
-    da_peaks_adata.uns["elapsed_time"] = time.time() - start_time
+    da_peaks_adata.uns["elapsed_time"] = time.perf_counter() - start_time
 
     adata.uns["step"] = 1
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/tool/_algorithm_.py

@@ -455,11 +455,11 @@ def semi_mutual_knn_weight(
         del data
         np.fill_diagonal(new_data, 0)
 
-    def _knn_k_(_mat: matrix_data, k: int):
+    def _knn_k_(_mat: matrix_data, k: int, info: str = "LOG"):
         n_rows = _mat.shape[0]
         adj = sparse.lil_matrix((n_rows, n_rows), dtype=np.int8)
 
-        ul.log(__name__).info("Calculate the k-nearest neighbors of each node.")
+        ul.log(__name__).info(f"Calculate the k-nearest neighbors of each node. ({info})")
 
         for i in tqdm(range(n_rows)):
             row = np.array(_mat[i]).ravel()
@@ -477,7 +477,7 @@ def semi_mutual_knn_weight(
 
         return adj.tocsr()
 
-    def _knn(_mat: matrix_data, k: int) -> matrix_data:
+    def _knn(_mat: matrix_data, k: int, info: str = "LOG") -> matrix_data:
         """
         Return k-nearest-neighbor 0/1 adjacency matrix (int8 to save memory).
         Supports both sparse and dense inputs.
@@ -486,11 +486,11 @@ def semi_mutual_knn_weight(
         if sparse.issparse(_mat):
             # Sparse path: sort each row's data to find the k-th largest
             _mat = _mat.tocsr(copy=False)
-            return _knn_k_(_mat, k)
+            return _knn_k_(_mat, k, info)
         else:
 
             if is_for:
-                return _knn_k_(_mat, k)
+                return _knn_k_(_mat, k, info)
             else:
                 # Dense path: vectorized thresholding
                 kth_val = np.sort(_mat, axis=1)[:, -(k + 1)]
@@ -499,12 +499,12 @@ def semi_mutual_knn_weight(
                 return adj
 
     # Compute adjacency matrices for AND/OR logic
-    adj_and = _knn(new_data, neighbors)
+    adj_and = _knn(new_data, neighbors, "AND")
 
     if neighbors == or_neighbors:
         adj_or = adj_and
     else:
-        adj_or = _knn(new_data, or_neighbors)
+        adj_or = _knn(new_data, or_neighbors, "OR")
 
     # Symmetrize
     if sparse.issparse(adj_and):
@@ -519,7 +519,12 @@ def semi_mutual_knn_weight(
 
     # Ensure full connectivity if required
     if is_mknn_fully_connected:
-        adj_1nn = _knn(new_data, 1)
+        adj_1nn = _knn(new_data, 1, "ONE")
+
+        if sparse.issparse(adj_1nn):
+            adj_1nn = adj_1nn.maximum(adj_1nn.T)
+        else:
+            adj_1nn = np.maximum(adj_1nn, adj_1nn.T)
 
         if sparse.issparse(adj_and):
             adj_and = adj_and.maximum(adj_1nn)
@@ -908,7 +913,7 @@ def overlap_sum(regions: AnnData, variants: dict, trait_info: DataFrame, n_jobs:
     :return: overlap data
     """
 
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     # Unique feature set
     label_all = regions.var.index.tolist()
@@ -1020,7 +1025,7 @@ def overlap_sum(regions: AnnData, variants: dict, trait_info: DataFrame, n_jobs:
 
     overlap_adata = AnnData(overlap_sparse, var=trait_info, obs=regions.var)
     overlap_adata.uns["is_overlap"] = True
-    overlap_adata.uns["elapsed_time"] = time.time() - start_time
+    overlap_adata.uns["elapsed_time"] = time.perf_counter() - start_time
 
     return overlap_adata
 
@@ -1124,7 +1129,7 @@ def calculate_init_score_weight(
     :return: Initial TRS with weight.
     """
 
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     if "is_overlap" not in overlap_adata.uns:
         ul.log(__name__).warning(
@@ -1253,7 +1258,7 @@ def calculate_init_score_weight(
     del _init_trs_ncw_, _cell_type_weight_
 
     init_trs_adata.uns["is_sample"] = is_simple
-    init_trs_adata.uns["elapsed_time"] = time.time() - start_time
+    init_trs_adata.uns["elapsed_time"] = time.perf_counter() - start_time
     return init_trs_adata
 
 
@@ -1309,7 +1314,7 @@ def obtain_cell_cell_network(
     :return: Cell similarity data.
     """
 
-    start_time = time.time()
+    start_time = time.perf_counter()
 
     from sklearn.metrics.pairwise import laplacian_kernel, rbf_kernel
 
@@ -1368,7 +1373,7 @@ def obtain_cell_cell_network(
     if not is_simple:
         cc_data.layers["cell_mutual_knn"] = to_sparse(cell_mutual_knn)
 
-    cc_data.uns["elapsed_time"] = time.time() - start_time
+    cc_data.uns["elapsed_time"] = time.perf_counter() - start_time
 
     return cc_data
 

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/tool/_random_walk_.py

@@ -282,7 +282,7 @@ class RandomWalk:
         """
         ul.log(__name__).info("Random walk with weighted seed cells.")
 
-        start_time = time.time()
+        start_time = time.perf_counter()
 
         # judge length
         if cc_adata.shape[0] != init_status.shape[0]:
@@ -459,7 +459,7 @@ class RandomWalk:
         del self.cell_affinity
         del init_status
 
-        self.elapsed_time = time.time() - start_time
+        self.elapsed_time = time.perf_counter() - start_time
 
     def _random_walk_(
         self,

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/util/__init__.py

@@ -29,6 +29,7 @@ from ._constant_ import (
 from ._core_ import (
     file_method,
     log,
+    track_with_memory,
     to_dense,
     to_sparse,
     sum_min_max,
@@ -58,6 +59,7 @@ __all__ = [
     "log_file_path",
     "file_method",
     "log",
+    "track_with_memory",
     "path",
     "plot_color_types",
     "sparse_array",

{sciv-0.0.100 → sciv-0.0.102}/src/sciv/util/_core_.py

@@ -4,7 +4,11 @@ import math
 import os
 import random
 import string
-from typing import Tuple, Union, Literal
+import threading
+import time
+from functools import wraps
+from typing import Tuple, Union, Literal, Callable, Any
+import psutil
 
 import numpy as np
 import pandas as pd
@@ -35,6 +39,71 @@ def log(name: str = None) -> Logger:
     return Logger(name, log_path=os.path.join(ul.log_file_path, name), is_form_file=ul.is_form_log_file)
 
 
+def track_with_memory(is_monitor: bool = False , interval: float = 60) -> Callable:
+    """
+    Decorator: Records memory usage at fixed intervals during function execution and returns the result, elapsed time, and memory list.
+
+    Parameters
+    ----------
+    is_monitor : bool, optional
+        Whether to enable memory monitoring, default is False.
+    interval : float, optional
+        Sampling interval (seconds), default is 60 seconds.
+
+    Returns
+    -------
+    Callable
+        Decorator function; when the wrapped function is called, it returns a dictionary containing:
+        - 'result': the original function's return value
+        - 'time': function execution time (seconds) if is_monitor is True, otherwise None.
+        - 'memory': list of sampled memory usage (bytes) if is_monitor is True, otherwise None.
+    """
+
+    def decorator(func) -> Callable:
+        @wraps(func)
+        def wrapper(*args, **kwargs) -> Union[Any, dict]:
+
+            if not is_monitor:
+                return func(*args, **kwargs)
+
+            process = psutil.Process(os.getpid())
+
+            stop_monitor = False
+
+            mem_list = []
+
+            def monitor():
+                nonlocal stop_monitor
+
+                while not stop_monitor:
+                    current_mem = process.memory_info().rss
+                    mem_list.append(current_mem)
+
+                    time.sleep(interval)
+
+            t = threading.Thread(target=monitor, daemon=True)
+            t.start()
+
+            start_time = time.perf_counter()
+            _result_ = func(*args, **kwargs)
+            end_time = time.perf_counter()
+
+            stop_monitor = True
+            t.join()
+
+            exec_time = end_time - start_time
+
+            return {
+                'result': _result_,
+                'time': exec_time,
+                'memory': mem_list
+            }
+
+        return wrapper
+
+    return decorator
+
+
 def to_dense(sm: matrix_data, is_array: bool = False) -> dense_data:
     """
     Convert sparse matrix to dense matrix