std::stable_sort 与sort 区别

#include <algorithm> #include <vector> #include <iostream> bool compare(int a, int b) { return a > b; // 降序排列 } int main() { std::vector<int> v = {5, 2, 3, 2, 1}; std::stable_sort(v.begin(), v.end(), compare); for (int x : v) std::cout << x << " "; }

首先，我得回忆一下std::stable_sort的基本用法和与std::sort的区别。记得stable_sort是稳定排序，相同元素的相对位置不变，而sort可能不保证这一点。用户想要的是自定义比较函数的示例，所以重点应该放在如何正确...

#include <algorithm> #include <vector> #include <iostream> bool compare(int a, int b) { return a > b; // 降序排列 } int main() { std::vector<int> v = {5, 2, 3, 2, 1}; std::stable_sort(v.begin(), v.end(), compare); // 输出：5 3 2 2 1 for (int x : v) std::cout << x << " "; }这段代码如果用数组实现有什么注意事项吗

标准库的std::sort通常是不稳定排序，而std::stable_sort是稳定排序。所以如果用户需要稳定排序，应该使用stable_sort函数，并自定义比较函数来实现降序。另外，比较函数的编写要注意正确性。比如，降序排序的比较...

std stable_sort源码

### C++ std::stable_sort 源码解析 #### 函数签名 std::stable_sort 是 C++ 标准模板库 (STL) 中的一个算法，用于执行稳定排序操作。该函数有两种重载形式： cpp template void stable_sort(RandomIt ...

import os import numpy as np import matplotlib.pyplot as plt import re from matplotlib.ticker import MaxNLocator # 解决中文显示问题 plt.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei', 'WenQuanYi Micro Hei'] plt.rcParams['axes.unicode_minus'] = False def natural_sort_key(s): """自然排序算法：确保文件名按数字顺序排列""" return [int(text) if text.isdigit() else text.lower() for text in re.split(r'(\d+)', s)] def find_stable_intervals(counts, min_window=300, max_window=2000, std_threshold=10.0, merge_gap=300, min_length=500): """ 改进版稳定区间检测：使用标准差作为稳定性指标 :param counts: 预测框数量列表 :param min_window: 最小窗口尺寸 :param max_window: 最大窗口尺寸 :param std_threshold: 标准差阈值（波动范围） :param merge_gap: 相邻区间合并的最大间隔 :param min_length: 最小有效区间长度 :return: 优化后的稳定区间列表 """ n = len(counts) if n == 0: return [] # 1. 自适应窗口机制 window_size = min(max_window, max(min_window, n // 10)) step_size = max(1, window_size // 2) # 50%重叠滑动 # 2. 初始检测稳定区间（使用标准差） base_intervals = [] for i in range(0, n - window_size + 1, step_size): window = counts[i:i + window_size] if len(window) < 2: # 至少需要2个点计算标准差 continue # 计算标准差作为稳定性指标 std_dev = np.std(window) if std_dev < std_threshold: base_intervals.append((i, i + window_size - 1)) # 如果没有检测到任何区间，直接返回 if not base_intervals: return [] # 3. 合并相邻平稳段 base_intervals.sort(key=lambda x: x[0]) # 确保按起始索引排序 merged_intervals = [] current_start, current_end = base_intervals[0] for start, end in base_intervals[1:]: if start - current_end <= merge_gap: # 间隔小于合并阈值 current_end = max(current_end, end) # 扩展当前区间 else: merged_intervals.append((current_start, current_end)) current_start, current_end = start, end merged_intervals.append((current_start, current_end)) # 4. 过滤短时伪平稳段 final_intervals = [ (start, end) for start, end in merged_intervals if (end - start + 1) >= min_length # 区间长度包含两端点 ] return final_intervals def plot_box_count_trend_with_stable_intervals(file_list, box_counts, stable_intervals, output_path): """ 绘制预测框数量变化趋势图并标记稳定区间 :param file_list: 文件名列表 :param box_counts: 预测框数量列表 :param stable_intervals: 稳定区间列表 :param output_path: 输出图片路径 """ plt.figure(figsize=(20, 10)) # 绘制整体趋势 plt.plot(file_list, box_counts, 'b-', linewidth=1.5, label='预测框数量') # 标记稳定区间 for i, (start, end) in enumerate(stable_intervals): interval_files = file_list[start:end + 1] interval_counts = box_counts[start:end + 1] if not interval_counts: # 确保区间有效 continue # 计算区间统计量 avg_count = np.mean(interval_counts) min_count = np.min(interval_counts) max_count = np.max(interval_counts) std_dev = np.std(interval_counts) # 绘制稳定区间 plt.fill_between(interval_files, min_count, max_count, color='green', alpha=0.2, label=f'稳定区间{i + 1}' if i == 0 else "") # 添加区间标注 mid_idx = start + (end - start) // 2 plt.annotate(f"区间{i + 1}: {start + 1}-{end + 1}\n均值: {avg_count:.1f}±{std_dev:.1f}", (file_list[mid_idx], avg_count), xytext=(0, 20), textcoords='offset points', ha='center', fontsize=10, bbox=dict(boxstyle="round,pad=0.3", fc="yellow", alpha=0.7)) # 设置图表属性 plt.title('预测框数量变化趋势及稳定区间分析', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel('预测框数量', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) plt.legend(loc='upper right') # 添加统计信息 stats_text = f"总文件数: {len(file_list)}\n稳定区间数: {len(stable_intervals)}" plt.figtext(0.95, 0.95, stats_text, ha='right', va='top', bbox=dict(facecolor='white', alpha=0.8), fontsize=12) # 限制X轴刻度数量 plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() # 配置路径 label_dir = "D:/630-3-label-combine" # 替换为您的标签文件夹路径 output_dir = "D:/630-report" # 输出目录 os.makedirs(output_dir, exist_ok=True) # 获取文件列表并按自然顺序排序 file_list = [f for f in os.listdir(label_dir) if f.endswith(".txt")] file_list.sort(key=natural_sort_key) # 提取文件名（不含扩展名） file_names = [os.path.splitext(f)[0] for f in file_list] # 统计每个文件的预测框数量 box_counts = [] for file in file_list: file_path = os.path.join(label_dir, file) count = 0 with open(file_path, 'r') as f: for line in f: if line.strip(): # 非空行 count += 1 box_counts.append(count) # 计算整体统计数据 total_mean = np.mean(box_counts) total_std = np.std(box_counts) # 找出稳定区间（使用标准差作为指标） stable_intervals = find_stable_intervals( box_counts, min_window=300, # 最小检测窗口 max_window=2000, # 最大检测窗口 std_threshold=total_std * 0.5, # 基于整体标准差设置阈值 merge_gap=300, # 合并最大间隔 min_length=500 # 最小有效长度 ) # 生成结果图片 output_path = os.path.join(output_dir, "box_count_stable_intervals_std.png") plot_box_count_trend_with_stable_intervals(file_names, box_counts, stable_intervals, output_path) # 输出详细结果 print(f"分析完成! 共处理 {len(file_list)} 个文件") print(f"整体平均框数: {total_mean:.2f} ± {total_std:.2f}") print(f"发现 {len(stable_intervals)} 个稳定区间:") for i, (start, end) in enumerate(stable_intervals): interval_counts = box_counts[start:end + 1] avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) print(f"区间{i + 1}:") print(f" - 文件范围: {start + 1}-{end + 1} (共{end - start + 1}个文件)") print(f" - 平均框数: {avg_count:.2f} ± {std_dev:.2f}") print(f" - 最小值: {min(interval_counts)}, 最大值: {max(interval_counts)}") print(f"结果图片已保存至: {output_path}") # 保存区间信息到文本文件 interval_info_path = os.path.join(output_dir, "stable_intervals_report_std.txt") with open(interval_info_path, 'w') as f: f.write(f"稳定区间分析报告（标准差指标）\n") f.write(f"总文件数: {len(file_list)}\n") f.write(f"整体平均框数: {total_mean:.2f} ± {total_std:.2f}\n") f.write(f"稳定区间数: {len(stable_intervals)}\n\n") for i, (start, end) in enumerate(stable_intervals): interval_counts = box_counts[start:end + 1] avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) f.write(f"区间 {i + 1}:\n") f.write(f" 起始文件索引: {start + 1} ({file_names[start]})\n") f.write(f" 结束文件索引: {end + 1} ({file_names[end]})\n") f.write(f" 文件数量: {end - start + 1}\n") f.write(f" 平均预测框数: {avg_count:.2f} ± {std_dev:.2f}\n") f.write(f" 最小值: {min(interval_counts)}, 最大值: {max(interval_counts)}\n") f.write("-" * 50 + "\n") print(f"详细区间报告已保存至: {interval_info_path}") 将代码中的预测框的数量改成txt文件中的x坐标的均值，y坐标的均值，分别生成图片和文件

def plot_coordinate_trend(file_list, x_means, y_means, stable_intervals_x, stable_intervals_y, output_dir, coord_type): """ 绘制坐标均值变化趋势图并标记稳定区间 :param file_list: 文件名列表 :param...

在vs2015 c++ .h中加入这段代码会报重定义 namespace cv_dnn { namespace { template <typename T> static inline bool SortScorePairDescend(const std::pair<float, T>& pair1, const std::pair<float, T>& pair2) { return pair1.first > pair2.first; } } // namespace inline void GetMaxScoreIndex(const std::vector<float>& scores, const float threshold, const int top_k, std::vector<std::pair<float, int> >& score_index_vec) { for (size_t i = 0; i < scores.size(); ++i) { if (scores[i] > threshold) { score_index_vec.push_back(std::make_pair(scores[i], i)); } } std::stable_sort(score_index_vec.begin(), score_index_vec.end(), SortScorePairDescend<int>); if (top_k > 0 && top_k < (int)score_index_vec.size()) { score_index_vec.resize(top_k); } } template <typename BoxType> inline void NMSFast_(const std::vector<BoxType>& bboxes, const std::vector<float>& scores, const float score_threshold, const float nms_threshold, const float eta, const int top_k, std::vector<int>& indices, float(computeOverlap)(const BoxType&, const BoxType&)) { CV_Assert(bboxes.size() == scores.size()); std::vector<std::pair<float, int> > score_index_vec; GetMaxScoreIndex(scores, score_threshold, top_k, score_index_vec); // Do nms. float adaptive_threshold = nms_threshold; indices.clear(); for (size_t i = 0; i < score_index_vec.size(); ++i) { const int idx = score_index_vec[i].second; bool keep = true; for (int k = 0; k < (int)indices.size() && keep; ++k) { const int kept_idx = indices[k]; float overlap = computeOverlap(bboxes[idx], bboxes[kept_idx]); keep = overlap <= adaptive_threshold; } if (keep) indices.push_back(idx); if (keep && eta < 1 && adaptive_threshold > 0.5) { adaptive_threshold = eta; } } } // copied from opencv 3.4, not exist in 3.0 template<typename Tp> static inline double jaccardDistance_(const Rect_<Tp>& a, const Rect<_Tp>& b) { Tp Aa = a.area(); Tp Ab = b.area(); if ((Aa + Ab) <= std::numeric_limits<Tp>::epsilon()) { // jaccard_index = 1 -> distance = 0 return 0.0; } double Aab = (a & b).area(); // distance = 1 - jaccard_index return 1.0 - Aab / (Aa + Ab - Aab); } template <typename T> static inline float rectOverlap(const T& a, const T& b) { return 1.f - static_cast<float>(jaccardDistance(a, b)); } void NMSBoxes(const std::vector<Rect>& bboxes, const std::vector<float>& scores, const float score_threshold, const float nms_threshold, std::vector<int>& indices, const float eta = 1, const int top_k = 0) { NMSFast(bboxes, scores, score_threshold, nms_threshold, eta, top_k, indices, rectOverlap); } }

在加入这段代码时，如果在同一个命名空间内有相同的函数或变量，则会报重定义的错误。如果你确定已经检查过代码，在其他地方没有定义相同的函数或变量，可以尝试将代码放在一个新的命名空间中。...

stable_sort

与 std::sort 不同的是，std::stable_sort 会保留相同元素的相对位置，即如果序列中有两个元素值相同，那么排序后它们的相对位置不会改变。 std::stable_sort 的时间复杂度为 $O(NlogN)$，其中 $N$ 是序列中...

C++ sort stable_sort

### C++ 中 std::sort 和 std::stable_sort 的区别及用法 #### 函数签名两个函数都位于 <algorithm> 头文件中，用于对容器中的元素进行排序。对于 std::sort： cpp template void sort(RandomIt ...

c++ stable_sort 和 sort

C++中有两个常用的排序函数：std::sort和std::stable_sort。两者都可以用于对容器中的元素进行排序，但在某些情况下它们的行为稍有不同。 std::sort函数使用快速排序算法进行排序，它具有较快的执行速度，但...

import os import numpy as np import matplotlib.pyplot as plt import re from matplotlib.ticker import MaxNLocator from scipy.stats import linregress import matplotlib.patches as mpatches # 解决中文显示问题 plt.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei', 'WenQuanYi Micro Hei'] plt.rcParams['axes.unicode_minus'] = False def natural_sort_key(s): """自然排序算法：确保文件名按数字顺序排列""" return [int(text) if text.isdigit() else text.lower() for text in re.split(r'(\d+)', s)] def calculate_coordinate_means(file_path): """ 计算文件中所有边界框的x坐标均值和y坐标均值 :param file_path: 标签文件路径 :return: (x_mean, y_mean) """ x_coords = [] y_coords = [] with open(file_path, 'r') as f: for line in f: if line.strip(): # 非空行 parts = line.split() if len(parts) >= 5: # 确保有足够的坐标信息 try: # YOLO格式：class_id x_center y_center width height x_center = float(parts[1]) y_center = float(parts[2]) x_coords.append(x_center) y_coords.append(y_center) except ValueError: continue # 计算均值，如果文件中没有目标框，返回(0,0) x_mean = np.mean(x_coords) if x_coords else 0.0 y_mean = np.mean(y_coords) if y_coords else 0.0 return x_mean, y_mean def find_stable_intervals(values, method='std', min_window=300, max_window=2000, threshold=0.5, merge_gap=300, min_length=500, return_base=False): """ 改进版稳定区间检测：支持三种不同指标 :param values: 坐标均值列表（原始数据） :param method: 检测方法 ('std', 'zscore', 'slope') :param min_window: 最小窗口尺寸 :param max_window: 最大窗口尺寸 :param threshold: 阈值（基于整体统计量） :param merge_gap: 相邻区间合并的最大间隔 :param min_length: 最小有效区间长度 :param return_base: 是否返回基础区间（合并前） :return: 优化后的稳定区间列表 """ n = len(values) if n == 0: return [] if not return_base else ([], [], []) # 计算整体统计量 total_mean = np.mean(values) total_std = np.std(values) # 1. 改进的窗口扩展检测机制 base_intervals = [] i = 0 # 当前检测起始点 while i < n: # 跳过已检测区域 if base_intervals and i <= base_intervals[-1][1]: i = base_intervals[-1][1] + 1 continue # 确保有足够的数据点 if i + min_window > n: break window_size = min_window stable_found = False # 检查当前窗口是否稳定 window = values[i:i + window_size] if len(window) < 2: i += 1 continue # 检测稳定性 if method == 'std': std_dev = np.std(window) stable = std_dev < threshold elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean(z_scores) stable = avg_zscore < threshold else: stable = True # 所有值相同 elif method == 'slope': x = np.arange(len(window)) slope, _, _, _, _ = linregress(x, window) stable = abs(slope) < threshold # 如果稳定，则尝试扩展窗口 if stable: stable_found = True start_idx = i end_idx = i + window_size - 1 # 尝试扩展窗口 while end_idx < n - 1 and window_size < max_window: # 扩展一个点 window_size += 1 new_end = i + window_size - 1 if new_end >= n: break # 获取新窗口 window = values[i:new_end + 1] if len(window) < 2: break # 检测新窗口的稳定性 if method == 'std': std_dev = np.std(window) current_stable = std_dev < threshold elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean(z_scores) current_stable = avg_zscore < threshold else: current_stable = True elif method == 'slope': x = np.arange(len(window)) slope, _, _, _, _ = linregress(x, window) current_stable = abs(slope) < threshold # 如果仍然稳定，更新结束索引 if current_stable: end_idx = new_end else: # 不稳定，回退到上一个稳定点 window_size -= 1 break # 添加找到的稳定区间 base_intervals.append((start_idx, end_idx)) i = end_idx + 1 # 从结束点后开始新检测 else: # 不稳定，移动到下一个点 i += 1 # 如果没有检测到任何区间，直接返回 if not base_intervals: if return_base: return [], [], [] return [] # 2. 合并相邻平稳段 base_intervals.sort(key=lambda x: x[0]) # 确保按起始索引排序 merged_intervals = [] current_start, current_end = base_intervals[0] for start, end in base_intervals[1:]: if start - current_end <= merge_gap: # 间隔小于合并阈值 current_end = max(current_end, end) # 扩展当前区间 else: merged_intervals.append((current_start, current_end)) current_start, current_end = start, end merged_intervals.append((current_start, current_end)) # 3. 过滤短时伪平稳段 final_intervals = [ (start, end) for start, end in merged_intervals if (end - start + 1) >= min_length # 区间长度包含两端点 ] if return_base: return base_intervals, merged_intervals, final_intervals return final_intervals def plot_coordinate_trend(file_list, values, stable_intervals, output_path, title_suffix="", method_name="标准差", coordinate_type="x坐标", is_base=False): """ 绘制坐标均值变化趋势图并标记稳定区间 :param is_base: 是否为基本区间图（合并前） :param coordinate_type: 坐标类型 ("x坐标" 或 "y坐标") """ plt.figure(figsize=(20, 10)) # 绘制整体趋势（原始数据） plt.plot(file_list, values, 'b-', linewidth=1.5, label=f'{coordinate_type}均值') # 计算全局最小值和最大值（用于颜色块填充） global_min = min(values) - 0.01 global_max = max(values) + 0.01 # 根据方法名称设置颜色 method_colors = { '标准差方法': 'green', 'Z-score方法': 'purple', '趋势斜率方法': 'orange' } # 获取当前方法的颜色 fill_color = method_colors.get(method_name, 'green') # 默认绿色 # 标记稳定区间 for i, (start, end) in enumerate(stable_intervals): interval_files = file_list[start:end + 1] if not interval_files: continue # 绘制稳定区间 - 使用对应方法的颜色 if is_base: # 基本区间图：颜色块只覆盖数据范围（避免叠加） interval_values = values[start:end + 1] min_val = min(interval_values) if interval_values else 0 max_val = max(interval_values) if interval_values else 0 plt.fill_between(interval_files, min_val, max_val, color=fill_color, alpha=0.3, zorder=0, label=f'{method_name}区间' if i == 0 else "") else: # 稳定区间图：颜色块顶到图表边缘 plt.fill_between(interval_files, global_min, global_max, color=fill_color, alpha=0.3, zorder=0, label=f'{method_name}区间' if i == 0 else "") # 如果是基本区间图（合并前），不添加标注，避免过多标注 if not is_base: # 添加区间标注 interval_values = values[start:end + 1] if interval_values: avg_val = np.mean(interval_values) std_dev = np.std(interval_values) mid_idx = start + (end - start) // 2 if mid_idx < len(file_list): plt.annotate(f"区间{i + 1}: {start + 1}-{end + 1}\n均值: {avg_val:.4f}±{std_dev:.4f}", (file_list[mid_idx], avg_val), xytext=(0, 20), textcoords='offset points', ha='center', fontsize=10, bbox=dict(boxstyle="round,pad=0.3", fc="yellow", alpha=0.7), zorder=10) # 设置图表属性 plot_type = "基础区间" if is_base else "稳定区间" plt.title(f'{coordinate_type}均值变化趋势 - {method_name}{title_suffix} ({plot_type})', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel(f'{coordinate_type}均值', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) plt.legend(loc='upper right') plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() def plot_combined_intervals(file_list, values, intervals_std, intervals_zscore, intervals_slope, output_path, coordinate_type="x坐标"): """ 绘制三种方法检测结果的合并图 :param coordinate_type: 坐标类型 ("x坐标" 或 "y坐标") """ plt.figure(figsize=(20, 10)) # 绘制整体趋势（原始数据） plt.plot(file_list, values, 'b-', linewidth=1.5, label=f'{coordinate_type}均值') # 计算全局最小值和最大值（用于颜色块填充） global_min = min(values) - 0.01 global_max = max(values) + 0.01 # 为每种方法定义不同的颜色和标签 method_colors = { '标准差方法': ('green', '标准差区间'), 'Z-score方法': ('purple', 'Z-score区间'), '趋势斜率方法': ('orange', '趋势斜率区间') } # 绘制标准差方法的区间 for i, (start, end) in enumerate(intervals_std): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['标准差方法'][0], alpha=0.3, zorder=0, label=method_colors['标准差方法'][1] if i == 0 else "") # 绘制Z-score方法的区间 for i, (start, end) in enumerate(intervals_zscore): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['Z-score方法'][0], alpha=0.3, zorder=0, label=method_colors['Z-score方法'][1] if i == 0 else "") # 绘制趋势斜率方法的区间 for i, (start, end) in enumerate(intervals_slope): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['趋势斜率方法'][0], alpha=0.3, zorder=0, label=method_colors['趋势斜率方法'][1] if i == 0 else "") # 设置图表属性 plt.title(f'{coordinate_type}均值变化趋势及稳定区间分析 - 三种方法合并', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel(f'{coordinate_type}均值', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) plt.legend(loc='upper right') plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() def plot_std_slope_intervals(file_list, values, intervals_std, intervals_slope, output_path, coordinate_type="x坐标"): """ 绘制标准差和趋势斜率最终稳定区间的二合一图 :param file_list: 文件名列表 :param values: 坐标均值列表 :param intervals_std: 标准差方法的稳定区间 :param intervals_slope: 趋势斜率方法的稳定区间 :param output_path: 输出图片路径 :param coordinate_type: 坐标类型 ("x坐标" 或 "y坐标") """ plt.figure(figsize=(20, 10)) # 绘制整体趋势（原始数据） plt.plot(file_list, values, 'b-', linewidth=1.5, label=f'{coordinate_type}均值') # 计算全局最小值和最大值（用于颜色块填充） global_min = min(values) - 0.01 global_max = max(values) + 0.01 # 定义每种方法的颜色和标签 method_colors = { '标准差方法': ('green', '标准差稳定区间'), '趋势斜率方法': ('orange', '趋势斜率稳定区间') } # 绘制标准差方法的稳定区间 for i, (start, end) in enumerate(intervals_std): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['标准差方法'][0], alpha=0.3, zorder=0, label=method_colors['标准差方法'][1] if i == 0 else "") # 添加区间标注 interval_values = values[start:end + 1] if interval_values: avg_val = np.mean(interval_values) std_dev = np.std(interval_values) mid_idx = start + (end - start) // 2 if mid_idx < len(file_list): plt.annotate(f"STD区间{i + 1}: {start + 1}-{end + 1}\n均值: {avg_val:.4f}±{std_dev:.4f}", (file_list[mid_idx], avg_val), xytext=(0, 20), textcoords='offset points', ha='center', fontsize=10, bbox=dict(boxstyle="round,pad=0.3", fc="yellow", alpha=0.7), zorder=10) # 绘制趋势斜率方法的稳定区间 for i, (start, end) in enumerate(intervals_slope): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['趋势斜率方法'][0], alpha=0.3, zorder=0, label=method_colors['趋势斜率方法'][1] if i == 0 else "") # 添加区间标注 interval_values = values[start:end + 1] if interval_values: avg_val = np.mean(interval_values) std_dev = np.std(interval_values) mid_idx = start + (end - start) // 2 if mid_idx < len(file_list): plt.annotate(f"SLOPE区间{i + 1}: {start + 1}-{end + 1}\n均值: {avg_val:.4f}", (file_list[mid_idx], avg_val), xytext=(0, -30), textcoords='offset points', ha='center', fontsize=10, bbox=dict(boxstyle="round,pad=0.3", fc="cyan", alpha=0.7), zorder=10) # 创建图例 std_patch = mpatches.Patch(color='green', alpha=0.3, label='标准差稳定区间') slope_patch = mpatches.Patch(color='orange', alpha=0.3, label='趋势斜率稳定区间') data_line, = plt.plot([], [], 'b-', linewidth=1.5, label=f'{coordinate_type}均值') # 设置图表属性 plt.title(f'{coordinate_type}均值变化趋势及稳定区间分析 - 标准差与趋势斜率方法合并', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel(f'{coordinate_type}均值', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) # 添加图例 plt.legend(handles=[data_line, std_patch, slope_patch], loc='upper right') plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() def save_interval_report(intervals, method_name, file_path, values, coordinate_type="x坐标", is_base=False, is_merged=False): """保存区间信息到文本文件""" interval_type = "基础区间" if is_base else "稳定区间" interval_type = "合并区间" if is_merged else interval_type with open(file_path, 'a') as f: f.write(f"\n{method_name} {interval_type}分析报告 ({coordinate_type})\n") f.write(f"{interval_type}数: {len(intervals)}\n") for i, (start, end) in enumerate(intervals): interval_values = values[start:end + 1] avg_val = np.mean(interval_values) std_dev = np.std(interval_values) cv = std_dev / avg_val if avg_val > 0 else 0 # 计算趋势斜率 x = np.arange(len(interval_values)) slope, _, _, _, _ = linregress(x, interval_values) f.write(f"\n区间 {i + 1}:\n") f.write(f" 起始文件索引: {start + 1} ({file_names[start]})\n") f.write(f" 结束文件索引: {end + 1} ({file_names[end]})\n") f.write(f" 文件数量: {end - start + 1}\n") f.write(f" 平均{coordinate_type}: {avg_val:.4f} ± {std_dev:.4f}\n") f.write(f" 变异系数: {cv:.4f}\n") f.write(f" 趋势斜率: {slope:.6f}\n") f.write(f" 最小值: {min(interval_values):.4f}, 最大值: {max(interval_values):.4f}\n") f.write("=" * 80 + "\n") def print_interval_info(intervals, method_name, values, coordinate_type="x坐标"): """打印区间信息到控制台""" print(f"\n{method_name}发现 {len(intervals)} 个稳定区间 ({coordinate_type}):") for i, (start, end) in enumerate(intervals): interval_values = values[start:end + 1] avg_val = np.mean(interval_values) std_dev = np.std(interval_values) cv = std_dev / avg_val if avg_val > 0 else 0 # 计算趋势斜率 x = np.arange(len(interval_values)) slope, _, _, _, _ = linregress(x, interval_values) print(f"区间{i + 1}:") print(f" - 文件范围: {start + 1}-{end + 1} (共{end - start + 1}个文件)") print(f" - 平均{coordinate_type}: {avg_val:.4f} ± {std_dev:.4f}") print(f" - 变异系数: {cv:.4f}") print(f" - 趋势斜率: {slope:.6f}") print(f" - 最小值: {min(interval_values):.4f}, 最大值: {max(interval_values):.4f}") def merge_intervals(intervals, merge_gap=300, min_length=500): """合并重叠或接近的区间""" if not intervals: return [] # 按起始索引排序 intervals.sort(key=lambda x: x[0]) merged = [] current_start, current_end = intervals[0] for start, end in intervals[1:]: if start - current_end <= merge_gap: # 间隔小于合并阈值 current_end = max(current_end, end) # 扩展当前区间 else: merged.append((current_start, current_end)) current_start, current_end = start, end merged.append((current_start, current_end)) # 过滤短区间 final_merged = [ (start, end) for start, end in merged if (end - start + 1) >= min_length ] return final_merged # ====================== 主程序 ====================== if name == "main": # 配置路径 label_dir = "E:/0706/0706-labels" # 替换为您的标签文件夹路径 output_dir = "E:/0706/0706-stable4/xy-mean" # 输出目录 os.makedirs(output_dir, exist_ok=True) # 获取文件列表并按自然顺序排序 file_list = [f for f in os.listdir(label_dir) if f.endswith(".txt")] file_list.sort(key=natural_sort_key) # 提取文件名（不含扩展名） file_names = [os.path.splitext(f)[0] for f in file_list] # 统计每个文件的x坐标均值和y坐标均值 x_means = [] y_means = [] for file in file_list: file_path = os.path.join(label_dir, file) x_mean, y_mean = calculate_coordinate_means(file_path) x_means.append(x_mean) y_means.append(y_mean) # 计算整体统计数据 x_total_mean = np.mean(x_means) x_total_std = np.std(x_means) y_total_mean = np.mean(y_means) y_total_std = np.std(y_means) # 使用三种不同方法找出x坐标均值的稳定区间 base_std_x, merged_std_x, intervals_std_x = find_stable_intervals( x_means, method='std', min_window=500, max_window=2000, threshold=0.1, # 标准差阈值 merge_gap=300, min_length=600, return_base=True ) base_zscore_x, merged_zscore_x, intervals_zscore_x = find_stable_intervals( x_means, method='zscore_avg', min_window=500, max_window=2000, threshold=0.6, merge_gap=300, min_length=600, return_base=True ) base_slope_x, merged_slope_x, intervals_slope_x = find_stable_intervals( x_means, method='slope', min_window=500, max_window=2000, threshold=0.00001, # 趋势斜率阈值 merge_gap=300, min_length=600, return_base=True ) # 使用三种不同方法找出y坐标均值的稳定区间 base_std_y, merged_std_y, intervals_std_y = find_stable_intervals( y_means, method='std', min_window=500, max_window=2000, threshold=0.1, # 标准差阈值 merge_gap=300, min_length=600, return_base=True ) base_zscore_y, merged_zscore_y, intervals_zscore_y = find_stable_intervals( y_means, method='zscore_avg', min_window=500, max_window=2000, threshold=0.6, merge_gap=300, min_length=600, return_base=True ) base_slope_y, merged_slope_y, intervals_slope_y = find_stable_intervals( y_means, method='slope', min_window=500, max_window=2000, threshold=0.00001, # 趋势斜率阈值 merge_gap=300, min_length=600, return_base=True ) # ====================== 生成图表 (x坐标) ====================== # 生成三种方法的最终结果图片 output_std_x = os.path.join(output_dir, "x_mean_stable_intervals_std.png") output_zscore_x = os.path.join(output_dir, "x_mean_stable_intervals_zscore.png") output_slope_x = os.path.join(output_dir, "x_mean_stable_intervals_slope.png") # 生成三种方法的基础区间图片（合并前） output_base_std_x = os.path.join(output_dir, "x_mean_base_intervals_std.png") output_base_zscore_x = os.path.join(output_dir, "x_mean_base_intervals_zscore.png") output_base_slope_x = os.path.join(output_dir, "x_mean_base_intervals_slope.png") # 生成三种方法的合并区间图片（合并后但未过滤） output_merged_std_x = os.path.join(output_dir, "x_mean_merged_intervals_std.png") output_merged_zscore_x = os.path.join(output_dir, "x_mean_merged_intervals_zscore.png") output_merged_slope_x = os.path.join(output_dir, "x_mean_merged_intervals_slope.png") # 生成合并图 output_combined_x = os.path.join(output_dir, "x_mean_stable_intervals_combined.png") # 生成标准差和趋势斜率方法稳定区间合并图 output_std_slope_x = os.path.join(output_dir, "x_mean_stable_intervals_std_slope.png") # 绘制最终结果图表 (x坐标) plot_coordinate_trend(file_names, x_means, intervals_std_x, output_std_x, method_name="标准差方法", coordinate_type="x坐标") plot_coordinate_trend(file_names, x_means, intervals_zscore_x, output_zscore_x, method_name="Z-score方法", coordinate_type="x坐标") plot_coordinate_trend(file_names, x_means, intervals_slope_x, output_slope_x, method_name="趋势斜率方法", coordinate_type="x坐标") # 绘制基础区间图（合并前） plot_coordinate_trend(file_names, x_means, base_std_x, output_base_std_x, method_name="标准差方法", coordinate_type="x坐标", is_base=True) plot_coordinate_trend(file_names, x_means, base_zscore_x, output_base_zscore_x, method_name="Z-score方法", coordinate_type="x坐标", is_base=True) plot_coordinate_trend(file_names, x_means, base_slope_x, output_base_slope_x, method_name="趋势斜率方法", coordinate_type="x坐标", is_base=True) # 绘制合并区间图（合并后但未过滤） plot_coordinate_trend(file_names, x_means, merged_std_x, output_merged_std_x, method_name="标准差方法", coordinate_type="x坐标") plot_coordinate_trend(file_names, x_means, merged_zscore_x, output_merged_zscore_x, method_name="Z-score方法", coordinate_type="x坐标") plot_coordinate_trend(file_names, x_means, merged_slope_x, output_merged_slope_x, method_name="趋势斜率方法", coordinate_type="x坐标") # 生成合并图 plot_combined_intervals(file_names, x_means, intervals_std_x, intervals_zscore_x, intervals_slope_x, output_combined_x, coordinate_type="x坐标") # 生成标准差和趋势斜率方法稳定区间合并图 plot_std_slope_intervals(file_names, x_means, intervals_std_x, intervals_slope_x, output_std_slope_x, coordinate_type="x坐标") # ====================== 生成图表 (y坐标) ====================== # 生成三种方法的最终结果图片 output_std_y = os.path.join(output_dir, "y_mean_stable_intervals_std.png") output_zscore_y = os.path.join(output_dir, "y_mean_stable_intervals_zscore.png") output_slope_y = os.path.join(output_dir, "y_mean_stable_intervals_slope.png") # 生成三种方法的基础区间图片（合并前） output_base_std_y = os.path.join(output_dir, "y_mean_base_intervals_std.png") output_base_zscore_y = os.path.join(output_dir, "y_mean_base_intervals_zscore.png") output_base_slope_y = os.path.join(output_dir, "y_mean_base_intervals_slope.png") # 生成三种方法的合并区间图片（合并后但未过滤） output_merged_std_y = os.path.join(output_dir, "y_mean_merged_intervals_std.png") output_merged_zscore_y = os.path.join(output_dir, "y_mean_merged_intervals_zscore.png") output_merged_slope_y = os.path.join(output_dir, "y_mean_merged_intervals_slope.png") # 生成合并图 output_combined_y = os.path.join(output_dir, "y_mean_stable_intervals_combined.png") # 生成标准差和趋势斜率方法稳定区间合并图 output_std_slope_y = os.path.join(output_dir, "y_mean_stable_intervals_std_slope.png") # 绘制最终结果图表 (y坐标) plot_coordinate_trend(file_names, y_means, intervals_std_y, output_std_y, method_name="标准差方法", coordinate_type="y坐标") plot_coordinate_trend(file_names, y_means, intervals_zscore_y, output_zscore_y, method_name="Z-score方法", coordinate_type="y坐标") plot_coordinate_trend(file_names, y_means, intervals_slope_y, output_slope_y, method_name="趋势斜率方法", coordinate_type="y坐标") # 绘制基础区间图（合并前） plot_coordinate_trend(file_names, y_means, base_std_y, output_base_std_y, method_name="标准差方法", coordinate_type="y坐标", is_base=True) plot_coordinate_trend(file_names, y_means, base_zscore_y, output_base_zscore_y, method_name="Z-score方法", coordinate_type="y坐标", is_base=True) plot_coordinate_trend(file_names, y_means, base_slope_y, output_base_slope_y, method_name="趋势斜率方法", coordinate_type="y坐标", is_base=True) # 绘制合并区间图（合并后但未过滤） plot_coordinate_trend(file_names, y_means, merged_std_y, output_merged_std_y, method_name="标准差方法", coordinate_type="y坐标") plot_coordinate_trend(file_names, y_means, merged_zscore_y, output_merged_zscore_y, method_name="Z-score方法", coordinate_type="y坐标") plot_coordinate_trend(file_names, y_means, merged_slope_y, output_merged_slope_y, method_name="趋势斜率方法", coordinate_type="y坐标") # 生成合并图 plot_combined_intervals(file_names, y_means, intervals_std_y, intervals_zscore_y, intervals_slope_y, output_combined_y, coordinate_type="y坐标") # 生成标准差和趋势斜率方法稳定区间合并图 plot_std_slope_intervals(file_names, y_means, intervals_std_y, intervals_slope_y, output_std_slope_y, coordinate_type="y坐标") 将代码中的x坐标均值，y坐标均值改为x坐标方差，y坐标方差

current_stable = std_dev elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean...

stable_sort函数

与sort函数不同的是，stable_sort能够保证相等元素的原始顺序不会被改变。使用stable_sort函数需要包含头文件，其函数原型如下： template void stable_sort(RandomIt first, RandomIt last); template, ...

import os import numpy as np import matplotlib.pyplot as plt import re from matplotlib.ticker import MaxNLocator from scipy.stats import linregress # 解决中文显示问题 plt.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei', 'WenQuanYi Micro Hei'] plt.rcParams['axes.unicode_minus'] = False def natural_sort_key(s): """自然排序算法：确保文件名按数字顺序排列""" return [int(text) if text.isdigit() else text.lower() for text in re.split(r'(\d+)', s)] def find_stable_intervals(counts, method='std', min_window=300, max_window=2000, threshold=0.5, merge_gap=300, min_length=500, return_base=False): """ 改进版稳定区间检测：支持三种不同指标 :param counts: 预测框数量列表（原始数据） :param method: 检测方法 ('std', 'zscore', 'slope') :param min_window: 最小窗口尺寸 :param max_window: 最大窗口尺寸 :param threshold: 阈值（基于整体统计量） :param merge_gap: 相邻区间合并的最大间隔 :param min_length: 最小有效区间长度 :param return_base: 是否返回基础区间（合并前） :return: 优化后的稳定区间列表（如果return_base=True，则返回(base_intervals, merged_intervals, final_intervals)） """ n = len(counts) if n == 0: return [] # 返回空列表 # 计算整体统计量（基于原始极好） total_mean = np.mean(counts) total_std = np.std(counts) # 1. 改进的窗口扩展检测机制 base_intervals = [] i = 0 # 当前检测起始点 while i < n: # 跳过已检测区域 if base_intervals and i <= base_intervals[-1][1]: i = base_intervals[-1][1] + 1 continue # 确保有足够的数据点 if i + min_window > n: break window_size = min_window stable_found = False # 检查当前窗口是否稳定 window = counts[i:i + window_size] if len(window) < 2: i += 1 continue # 检测稳定性 if method == 'std': std_dev = np.std(window) stable = std_dev < threshold elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean(z_scores) stable = avg_zscore < threshold else: stable = True # 所有值相同 elif method == 'slope': x = np.arange(len(window)) slope, _, _, _, _ = linregress(x, window) stable = abs(slope) < threshold # 如果稳定，则尝试扩展窗口 if stable: stable_found = True start_idx = i end_idx = i + window_size - 1 # 尝试扩展窗口 while end_idx < n - 1 and window_size < max_window: # 扩展一个点 window_size += 1 new_end = i + window_size - 1 if new_end >= n: break # 获取新窗口 window = counts[i:new_end + 1] if len(window) < 2: break # 检测新窗口的稳定性 if method == 'std': std_dev = np.std(window) current_stable = std_dev < threshold elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean(z_scores) current_stable = avg_zscore < threshold else: current_stable = True elif method == 'slope': x = np.arange(len(window)) slope, _, _, _, _ = linregress(x, window) current_stable = abs(slope) < threshold # 如果仍然稳定，更新结束索引 if current_stable: end_idx = new_end else: # 不稳定，回退到上一个稳定点 window_size -= 1 break # 添加找到的稳定区间 base_intervals.append((start_idx, end_idx)) i = end_idx + 1 # 从结束点后开始新检测 else: # 不稳定，移动到下一个点 i += 1 # 如果没有检测到任何区间，直接返回 if not base_intervals: if return_base: return [], [], [] return [] # 返回空列表 # 2. 合并相邻平稳段 base_intervals.sort(key=lambda x: x[0]) # 确保按起始索引排序 merged_intervals = [] if base_intervals: # 确保列表非空 current_start, current_end = base_intervals[0] for start, end in base_intervals[1:]: if start - current_end <= merge_gap: # 间隔小于合并阈值 current_end = max(current_end, end) # 扩展当前区间 else: merged_intervals.append((current_start, current_end)) current_start, current_end = start, end merged_intervals.append((current_start, current_end)) # 3. 过滤短时伪平稳段 final_intervals = [ (start, end) for start, end in merged_intervals if (end - start + 1) >= min_length # 区间长度包含两端点 ] if return_base: return base_intervals, merged_intervals, final_intervals return final_intervals def plot_box_count_trend(file_list, box_counts, stable_intervals, output_path, title_suffix="", method_name="标准差", is_base=False): """ 绘制预测框数量变化趋势图并标记稳定区间修改：根据方法名称设置对应颜色，与合并图一致 :param is_base: 是否为基本区间图（合并前） """ plt.figure(figsize=(20, 10)) # 绘制整体趋势（原始数据） plt.plot(file_list, box_counts, 'b-', linewidth=1.5, label='预测框数量') # 计算全局最小值和最大值（用于颜色块填充） global_min = min(box_counts) - 0.5 global_max = max(box_counts) + 0.5 # 根据方法名称设置颜色（与合并图保持一致） method_colors = { '标准差方法': 'green', 'Z-score方法': 'purple', '趋势斜率方法': 'orange' } # 获取当前方法的颜色 fill_color = method_colors.get(method_name, 'green') # 默认绿色 # 标记稳定区间 for i, (start, end) in enumerate(stable_intervals): interval_files = file_list[start:end + 1] if not interval_files: continue # 绘制稳定区间 - 使用对应方法的颜色 if is_base: # 基本区间图：颜色块只覆盖数据范围（避免叠加） interval_counts = box_counts[start:end + 1] min_count = min(interval_counts) if interval_counts else 0 max_count = max(interval_counts) if interval_counts else 0 plt.fill_between(interval_files, min_count, max_count, color=fill_color, alpha=0.3, zorder=0, label=f'{method_name}区间' if i == 0 else "") else: # 稳定区间图：颜色块顶到图表边缘 plt.fill_between(interval_files, global_min, global_max, color=fill_color, alpha=0.3, zorder=0, label=f'{method_name}区间' if i == 0 else "") # 如果是基本区间图（合并前），不添加标注，避免过多标注 if not is_base: # 添加区间标注 interval_counts = box_counts[start:end + 1] if interval_counts: avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) mid_idx = start + (end - start) // 2 if mid_idx < len(file_list): plt.annotate(f"区间{i + 1}: {start + 1}-{end + 1}\n均值: {avg_count:.1f}±{std_dev:.1f}", (file_list[mid_idx], avg_count), xytext=(0, 20), textcoords='offset points', ha='center', fontsize=10, bbox=dict(boxstyle="round,pad=0.3", fc="yellow", alpha=0.7), zorder=10) # 设置图表属性 plot_type = "基础区间" if is_base else "稳定区间" plt.title(f'预测框数量变化趋势 - {method_name}{title_suffix} ({plot_type})', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel('预测框数量', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) plt.legend(loc='upper right') plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() def plot_combined_intervals(file_list, box_counts, intervals_std, intervals_zscore, intervals_slope, output_path): """ 绘制三种方法检测结果的合并图 :param file_list: 文件名列表 :param box_counts: 原始预测框数量列表 :param intervals_std: 标准差方法检测的区间 :param intervals_zscore: Z-score方法检测的区间 :param intervals_slope: 趋势斜率方法检测的区间 :param output_path: 输出图片路径 """ plt.figure(figsize=(20, 10)) # 绘制整体趋势（原始数据） plt.plot(file_list, box_counts, 'b-', linewidth=1.5, label='预测框数量') # 计算全局最小值和最大值（用于颜色块填充） global_min = min(box_counts) - 0.5 global_max = max(box_counts) + 0.5 # 为每种方法定义不同的颜色和标签（与单独图表一致） method_colors = { '标准差方法': ('green', '标准差区间'), 'Z-score方法': ('purple', 'Z-score区间'), '趋势斜率方法': ('orange', '趋势斜率区间') } # 绘制标准差方法的区间 for i, (start, end) in enumerate(intervals_std): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['标准差方法'][0], alpha=0.3, zorder=0, label=method_colors['标准差方法'][1] if i == 0 else "") # 绘制Z-score方法的区间 for i, (start, end) in enumerate(intervals_zscore): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['Z-score方法'][0], alpha=0.3, zorder=0, label=method_colors['Z-score方法'][1] if i == 0 else "") # 绘制趋势斜率方法的区间 for i, (start, end) in enumerate(intervals_slope): interval_files = file_list[start:end + 1] if not interval_files: continue plt.fill_between(interval_files, global_min, global_max, color=method_colors['趋势斜率方法'][0], alpha=0.3, zorder=0, label=method_colors['趋势斜率方法'][1] if i == 0 else "") # 设置图表属性 plt.title('预测框数量变化趋势及稳定区间分析 - 三种方法合并', fontsize=18) plt.xlabel('图像文件名', fontsize=14) plt.ylabel('预测框数量', fontsize=14) plt.xticks(rotation=90, fontsize=7) plt.grid(True, linestyle='--', alpha=0.6) plt.legend(loc='upper right') plt.gca().xaxis.set_major_locator(MaxNLocator(20)) plt.tight_layout() plt.savefig(output_path, dpi=150, bbox_inches='tight') plt.close() # 配置路径 label_dir = "E:/0706/0706-labels" # 替换为您的标签文件夹路径 output_dir = "E:/0706/0706-stable3" # 输出目录 os.makedirs(output_dir, exist_ok=True) # 获取文件列表并按自然顺序排序 file_list = [f for f in os.listdir(label_dir) if f.endswith(".txt")] file_list.sort(key=natural_sort_key) # 提取文件名（不含扩展名） file_names = [os.path.splitext(f)[0] for f in file_list] # 统计每个文件的预测框数量 box_counts = [] for file in file_list: file_path = os.path.join(label_dir, file) count = 0 with open(file_path, 'r') as f: for line in f: if line.strip(): # 非空行 count += 1 box_counts.append(count) # 计算整体统计数据 total_mean = np.mean(box_counts) total_std = np.std(box_counts) # 使用三种不同方法找出稳定区间（直接使用原始数据） # 修改为返回基础区间、合并区间和最终区间 base_std, merged_std, intervals_std = find_stable_intervals( box_counts, method='std', min_window=500, max_window=2000, threshold=1.1, # 标准差阈值 merge_gap=300, min_length=600, return_base=True ) base_zscore, merged_zscore, intervals_zscore = find_stable_intervals( box_counts, method='zscore_avg', min_window=500, max_window=2000, threshold=0.75, merge_gap=300, min_length=600, return_base=True ) base_slope, merged_slope, intervals_slope = find_stable_intervals( box_counts, method='slope', min_window=500, max_window=2000, threshold=0.000025, # 趋势斜率阈值 merge_gap=300, min_length=600, return_base=True ) # 生成三种方法的最终结果图片 output_std = os.path.join(output_dir, "box_count_stable_intervals_std.png") output_zscore = os.path.join(output_dir, "box_count_stable_intervals_zscore.png") output_slope = os.path.join(output_dir, "box_count_stable_intervals_slope.png") # 生成三种方法的基础区间图片（合并前） output_base_std = os.path.join(output_dir, "box_count_base_intervals_std.png") output_base_zscore = os.path.join(output_dir, "box_count_base_intervals_zscore.png") output_base_slope = os.path.join(output_dir, "box_count_base_intervals_slope.png") # 生成三种方法的合并区间图片（合并后但未过滤） output_merged_std = os.path.join(output_dir, "box_count_merged_intervals_std.png") output_merged_zscore = os.path.join(output_dir, "box_count_merged_intervals_zscore.png") output_merged_slope = os.path.join(output_dir, "box_count_merged_intervals_slope.png") # 生成合并图 output_combined = os.path.join(output_dir, "box_count_stable_intervals_combined.png") # 绘制最终结果图表（使用统一的方法名称） plot_box_count_trend(file_names, box_counts, intervals_std, output_std, title_suffix="", method_name="标准差方法") plot_box_count_trend(file_names, box_counts, intervals_zscore, output_zscore, title_suffix="", method_name="Z-score方法") plot_box_count_trend(file_names, box_counts, intervals_slope, output_slope, title_suffix="", method_name="趋势斜率方法") # 绘制基础区间图（合并前） plot_box_count_trend(file_names, box_counts, base_std, output_base_std, title_suffix="", method_name="标准差方法", is_base=True) plot_box_count_trend(file_names, box_counts, base_zscore, output_base_zscore, title_suffix="", method_name="Z-score方法", is_base=True) plot_box_count_trend(file_names, box_counts, base_slope, output_base_slope, title_suffix="", method_name="趋势斜率方法", is_base=True) # 绘制合并区间图（合并后但未过滤） plot_box_count_trend(file_names, box_counts, merged_std, output_merged_std, title_suffix="", method_name="标准差方法") plot_box_count_trend(file_names, box_counts, merged_zscore, output_merged_zscore, title_suffix="", method_name="Z-score方法") plot_box_count_trend(file_names, box_counts, merged_slope, output_merged_slope, title_suffix="", method_name="趋势斜率方法") # 生成合并图 plot_combined_intervals(file_names, box_counts, intervals_std, intervals_zscore, intervals_slope, output_combined) # 输出详细结果 print(f"分析完成! 共处理 {len(file_list)} 个文件") print(f"整体平均框数: {total_mean:.2f} ± {total_std:.2f}") def print_interval_info(intervals, method_name): print(f"\n{method_name}发现 {len(intervals)} 个稳定区间:") for i, (start, end) in enumerate(intervals): interval_counts = box_counts[start:end + 1] avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) cv = std_dev / avg_count if avg_count > 0 else 0 # 计算趋势斜率（基于原始数据） x = np.arange(len(interval_counts)) slope, _, _, _, _ = linregress(x, interval_counts) print(f"区间{i + 1}:") print(f" - 文件范围: {start + 1}-{end + 1} (共{end - start + 1}个文件)") print(f" - 平均框数: {avg_count:.2f} ± {std_dev:.2f}") print(f" - 变异系数: {cv:.4f}") print(f" - 趋势斜率: {slope:.6f}") print(f" - 最小值: {min(interval_counts)}, 最大值: {max(interval_counts)}") print_interval_info(intervals_std, "标准差方法") print_interval_info(intervals_zscore, "Z-score方法") print_interval_info(intervals_slope, "趋势斜率方法") # 合并所有检测到的区间 all_intervals = intervals_std + intervals_zscore + intervals_slope def merge_intervals(intervals, merge_gap=300, min_length=500): """合并重叠或接近的区间""" if not intervals: return [] # 按起始索引排序 intervals.sort(key=lambda x: x[0]) merged = [] current_start, current_end = intervals[0] for start, end in intervals[1:]: if start - current_end <= merge_gap: # 间隔小于合并阈值 current_end = max(current_end, end) # 扩展当前区间 else: merged.append((current_start, current_end)) current_start, current_end = start, end merged.append((current_start, current_end)) # 过滤短区间 final_merged = [ (start, end) for start, end in merged if (end - start + 1) >= min_length ] return final_merged # 合并所有检测到的区间 merged_intervals = merge_intervals(all_intervals, merge_gap=300, min_length=500) # 保存区间信息到文本文件 def save_interval_report(intervals, method_name, file_path, is_base=False, is_merged=False): interval_type = "基础区间" if is_base else "稳定区间" interval_type = "合并区间" if is_merged else interval_type with open(file_path, 'a') as f: f.write(f"\n{method_name} {interval_type}分析报告\n") f.write(f"{interval_type}数: {len(intervals)}\n") for i, (start, end) in enumerate(intervals): interval_counts = box_counts[start:end + 1] avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) cv = std_dev / avg_count if avg_count > 0 else 0 # 计算趋势斜率 x = np.arange(len(interval_counts)) slope, _, _, _, _ = linregress(x, interval_counts) f.write(f"\n区间 {i + 1}:\n") f.write(f" 起始文件索引: {start + 1} ({file_names[start]})\n") f.write(f" 结束文件索引: {end + 1} ({file_names[end]})\n") f.write(f" 文件数量: {end - start + 1}\n") f.write(f" 平均预测框数: {avg_count:.2f} ± {std_dev:.2f}\n") f.write(f" 变异系数: {cv:.4f}\n") f.write(f" 趋势斜率: {slope:.6f}\n") f.write(f" 最小值: {min(interval_counts)}, 最大值: {max(interval_counts)}\n") f.write("=" * 80 + "\n") # 创建报告文件 interval_info_path = os.path.join(output_dir, "stable_intervals_report.txt") with open(interval_info_path, 'w') as f: f.write(f"稳定区间综合分析报告\n") f.write(f"总文件数: {len(file_list)}\n") f.write(f"整体平均框数: {total_mean:.2f} ± {total_std:.2f}\极好") f.write(f"数据范围: {min(box_counts)}-{max(box_counts)}\n") # 保存三种方法的区间报告 save_interval_report(intervals_std, "标准差方法", interval_info_path) save_interval_report(intervals_zscore, "Z-score方法", interval_info_path) save_interval_report(intervals_slope, "趋势斜率方法", interval_info_path) # 保存基础区间报告 save_interval_report(base_std, "标准差方法", interval_info_path, is_base=True) save_interval_report(base_zscore, "Z-score方法", interval_info_path, is_base=True) save_interval_report(base_slope, "趋势斜率方法", interval_info_path, is_base=True) # 保存合并区间报告（合并后但未过滤） save_interval_report(merged_std, "标准差方法", interval_info_path, is_merged=True) save_interval_report(merged_zscore, "Z-score方法", interval_info_path, is_merged=True) save_interval_report(merged_slope, "趋势斜率方法", interval_info_path, is_merged=True) # 保存合并后的区间报告 with open(interval_info_path, 'a') as f: f.write("\n\n=== 合并区间分析报告 ===\n") f.write("此部分展示三种方法检测到的所有稳定区间合并后的结果\n") f.write(f"合并后稳定区间数: {len(merged_intervals)}\n") for i, (start, end) in enumerate(merged_intervals): interval_counts = box_counts[start:end + 1] avg_count = np.mean(interval_counts) std_dev = np.std(interval_counts) cv = std_dev / avg_count if avg_count > 0 else 0 # 计算趋势斜率 x = np.arange(len(interval_counts)) slope, _, _, _, _ = linregress(x, interval_counts) # 检测此区间被哪些方法覆盖 covered_by = [] if any(start >= s and end <= e for s, e in intervals_std): covered_by.append("标准差") if any(start >= s and end <= e for s, e in intervals_zscore): covered_by.append("Z-score") if any(start >= s and end <= e for s, e in intervals_slope): covered_by.append("趋势斜率") f.write(f"\n合并区间 {i + 1}:\n") f.write(f" 起始文件索引: {start + 1} ({file_names[start]})\n") f.write(f" 结束文件索引: {end + 1} ({file_names[end]})\n") f.write(f" 文件数量: {end - start + 1}\n") f.write(f" 平均预测框数: {avg_count:.2f} ± {std_dev:.2f}\n") f.write(f" 最小值: {min(interval_counts)}, 最大值: {max(interval_counts)}\n") f.write(f" 覆盖方法: {', '.join(covered_by) if covered_by else '无'}\n") # 添加合并区间统计 total_covered_files = sum(end - start + 1 for start, end in merged_intervals) coverage_percentage = (total_covered_files / len(file_list)) * 100 f.write("\n合并区间统计:\n") f.write(f" 总覆盖文件数: {total_covered_files}/{len(file_list)} ({coverage_percentage:.2f}%)\n") f.write(f" 平均区间长度: {np.mean([end - start + 1 for start, end in merged_intervals]):.1f} 文件\n") f.write(f" 最长区间: {max([end - start + 1 for start, end in merged_intervals])} 文件\n") f.write(f" 最短区间: {min([end - start + 1 for start, end in merged_intervals])} 文件\n") print_interval_info(merged_intervals, "合并区间") print(f"\n结果图片已保存至: {output_dir}") print(f"详细区间报告已保存至: {interval_info_path}") 将代码中的预测框数量换成X坐标均值，y坐标均值

current_stable = std_dev elif method == 'zscore_avg': mean_val = np.mean(window) std_val = np.std(window) if std_val > 0: z_scores = np.abs((window - mean_val) / std_val) avg_zscore = np.mean...

std::stable_sort

std::stable_sort 是 C++ 标准库 <algorithm> 模块中的一个函数模板，用于对容器中的元素进行稳定排序。所谓稳定排序，就是对于相等的元素，其原有的相对顺序在排序后保持不变。这意味着如果有两个元素 a 和 b，...

GCC std::stable_sort() 解析

好的，我现在要帮用户解析GCC中的std::stable_sort()的实现。首先，我得回忆一下std::stable_sort的基本概念。用户可能已经知道这是一个稳定的排序算法，相等元素的相对顺序会被保留。但用户想了解的是GCC具体如何...

一个专为IntelliJIDEA和AndroidStudio开发者设计的全能格式转换与数据辅助工具_支持JSON_XML_HTML格式化_MD5生成_QRCODE二维码生成_B.zip

基于 Python 机器学习的简易天气预报系统实现

资源下载链接为： https://pan.quark.cn/s/46acb4885600 基于 Python 机器学习的简易天气预报系统实现（最新、最全版本！打开链接下载即可用！）

resalloc-selinux-5.9-1.el8.tar.gz

# 适用操作系统：Centos8 #Step1、解压 tar -zxvf xxx.el8.tar.gz #Step2、进入解压后的目录，执行安装 sudo rpm -ivh *.rpm

reptyr-0.7.0-1.el8.tar.gz

# 适用操作系统：Centos8 #Step1、解压 tar -zxvf xxx.el8.tar.gz #Step2、进入解压后的目录，执行安装 sudo rpm -ivh *.rpm

多线程编程的基本概念

线程总是在某个进程环境中创建，并且会在这个进程内部销毁。线程和进程的关系是:线程是属于进程的，线程运行在进程空间内，同一进程所产生的线程共享同一内存空间，当进程退出时该进程所产生的线程都会被强制退出并清除

vue+SpringBoot智能阅读推荐系统毕业论文.docx

std::stable_sort 与sort 区别

相关推荐

C++编程C++标准库排序算法详解：常用函数、自定义规则及性能优化方法汇总

STL_Sort 排序的资料

C++中sort函数的基础入门使用教程

#include <algorithm> #include <vector> #include <iostream> bool compare(int a, int b) { return a > b; // 降序排列 } int main() { std::vector<int> v = {5, 2, 3, 2, 1}; std::stable_sort(v.begin(), v.end(), compare); for (int x : v) std::cout << x << " "; }

std stable_sort源码

stable_sort

C++ sort stable_sort

c++ stable_sort 和 sort

stable_sort函数

std::stable_sort

GCC std::stable_sort() 解析

一个专为IntelliJIDEA和AndroidStudio开发者设计的全能格式转换与数据辅助工具_支持JSON_XML_HTML格式化_MD5生成_QRCODE二维码生成_B.zip

基于 Python 机器学习的简易天气预报系统实现

resalloc-selinux-5.9-1.el8.tar.gz

reptyr-0.7.0-1.el8.tar.gz

多线程编程的基本概念

vue+SpringBoot智能阅读推荐系统毕业论文.docx

大家在看

批量提取eml

AMIDE-开源

基于 ADS9110的隔离式数据采集 (DAQ) 系统方案（待编辑）-电路方案

自动化图书管理系统 v7.0

白盒测试基本路径自动生成工具制作文档附代码

最新推荐

一个专为IntelliJIDEA和AndroidStudio开发者设计的全能格式转换与数据辅助工具_支持JSON_XML_HTML格式化_MD5生成_QRCODE二维码生成_B.zip

基于 Python 机器学习的简易天气预报系统实现

一步到位：Blender Flamenco服务器安装及使用指南

Coze工作流高级技巧：源码挖掘与性能调优秘技

我现在通过订阅burger小车的odom话题得知小车的实际位置，在gazebo仿真当中我可以改表小车的初始位置，但是在实物仿真我该如何改变小车的初始位置呢

Pandas库在Python中的挑战性应用教程

深入理解Coze工作流：源码解析与优化之道

编码器机器人：新经济系统与Discord.js开发指南

【Coze工作流终极指南】：2025版保姆级教程启动