Python基于机器学习的分布式系统故障诊断系统源代码，分布式系统的故障数据进行分析，设计故障诊断模型，高效地分析并识别故障类别

import os import pandas as pd import numpy as np from django.http import JsonResponse, HttpResponse from django.shortcuts import render from django.views import View from django.conf import settings import sklearn.metrics import lightgbm as lgb import catboost as cbt import xgboost as xgb from statistics import mode import sklearn.metrics from cvxopt import matrix, solvers import json from sklearn.metrics import f1_score from collections import Counter # Create combination views here. # 将上传的文件名作为全局变量保存 train_file_name = '' predict_file_name = '' # 上传训练集，接收训练集 class Upload(View): """训练功能：上传训练集""" def get(self, request): return render(request, 'combination/combination-upload.html') def post(self, request): global train_file_name global predict_file_name train_csv_file = request.FILES.get('train_csv_file') predict_csv_file = request.FILES.get('predict_csv_file') # 这里记得加入判断传入文件格式为空，以及文件格式不对的判断 train_file_name = train_csv_file.name predict_file_name = predict_csv_file.name if predict_file_name.split('.')[1] != 'csv' and train_file_name.split('.')[1] != 'csv': return JsonResponse({'code': 0}) train_file_path = os.path.join(settings.MEDIA_ROOT, 'csv_dir\\' + train_file_name) predict_file_path = os.path.join(settings.MEDIA_ROOT, 'csv_dir\\' + predict_file_name) # 学到 with open(train_file_path, 'wb+') as f: for chunk in train_csv_file.chunks(): f.write(chunk) with open(predict_file_path, 'wb+') as f: for chunk in predict_csv_file.chunks(): f.write(chunk) return JsonResponse({'code': 1}) # 执行训练 def execute_combination(request): global train_file_name global predict_file_name # ============================ 数据预处理 ================================= # # 读入 CSV 文件 df = pd.read_csv(os.path.join(settings.MEDIA_ROOT, 'csv_dir\\' + train_file_name)) # df1 = pd.read_csv(os.path.join(settings.STATICFILES_DIRS[0], 'csv_dir\\' + 'validate_1000.csv')) df1 = pd.read_csv(os.path.join(settings.STATIC_ROOT, 'csv_dir\\' + 'validate_1000.csv')) df2 = pd.read_csv(os.path.join(settings.MEDIA_ROOT, 'csv_dir\\' + predict_file_name)) # 获取每列的平均值，并用该列的平均值填充空值和 NaN 值 df = df.apply(lambda x: x.fillna(x.mean()), axis=0) df1 = df1.apply(lambda x: x.fillna(x.mean()), axis=0) df2 = df2.apply(lambda x: x.fillna(x.mean()), axis=0) Xtrain = df.iloc[::, 1:108:] Ytrain = df.iloc[:, -1] Xva = df1.iloc[::, 1:108:] Yva = df1.iloc[:, -1] Xtest = df2.iloc[::, 1:108:] # 训练集的样本数量、特征数 sample_size_train = len(Xtrain) # 行，样本数量 feature_quantity_train = len(Xtrain.columns) # 列，特征数 # 测试集的样本数量、特征数 sample_size_test = len(Xtest) # 行，样本数量 feature_quantity_test = len(Xtest.columns) # 列，特征数 kmm = KMM(kernel_type='rbf', B=1) beta = kmm.fit(Xtrain, Xva) Xtrain = pd.DataFrame(Xtrain) Xs_new = beta * Xtrain # ==================================训练基础分类器==================================== # xg = xgb.XGBClassifier(seed=0) xg.fit(Xs_new, Ytrain) y_pred = xg.predict(Xva) xg_f1 = f1_score(Yva, y_pred, average=None) cb = cbt.CatBoostClassifier(eval_metric='AUC') cb.fit(Xs_new, Ytrain) y_pred = cb.predict(Xva) cb_f1 = f1_score(Yva, y_pred, average=None) lg = lgb.LGBMClassifier() lg.fit(Xs_new, Ytrain) y_pred = lg.predict(Xva) lg_f1 = f1_score(Yva, y_pred, average=None) # ==================================预测==================================== # model = [] for i in range(len(lg_f1)): if max(lg_f1[i], xg_f1[i], cb_f1[i]) == lg_f1[i]: model.append(lg) elif max(lg_f1[i], xg_f1[i], cb_f1[i]) == xg_f1[i]: model.append(xg) else: model.append(cb) yp = our_model(Xtest, m1=lg, m2=xg, m3=cb, model=model) # ==================================可视化==================================== # counter = Counter(yp) # 输出分类结果（JSON格式） unique = [] counts = [] for k, v in counter.items(): unique.append(k) counts.append(v) # 对输出结果排序 unique.sort() # 生成统计每种分类的数量 counts = [dict(counter)[k] for k in unique] # 转变为echarts可以接收的格式 pie_data = dict(counter) # 将int64类型的数据转换为int类型 df['sample_id'] = df['sample_id'].astype(int) # 提取sample_id和预测值列的数据 sample_ids = df['sample_id'].tolist() predictions = yp # 创建字典 output_dict = {str(sample_id): prediction for sample_id, prediction in zip(sample_ids, predictions)} # 指定保存JSON文件的路径和文件名 output_file = 'output_combination.json' # 将字典保存为JSON文件，指定ensure_ascii=False参数 with open(os.path.join(settings.MEDIA_ROOT, 'result\\' + output_file), 'w') as f: json.dump(output_dict, f, ensure_ascii=False) # 返回结果给可视化界面 context = { 'unique': unique, 'counts': counts, 'pie_data': pie_data, 'sample_size_train': sample_size_train, 'feature_quantity_train': feature_quantity_train, 'sample_size': sample_size_test, 'feature_quantity': feature_quantity_test, 'describe': '测试集', } response = render(request, 'combination/combination-visualization.html', context=context) combination_count = int(request.COOKIES.get('combination')) combination_count += 1 response.set_cookie('combination', combination_count) return response # 下载预测结果 def download(request): # 读取要下载的文件 filepath = os.path.join(settings.MEDIA_ROOT, 'result\\' + 'output_combination.json') with open(filepath, 'rb') as f: filedata = f.read() # 返回响应，弹出文件下载对话框 response = HttpResponse(filedata, content_type='application/json') response['Content-Disposition'] = f'attachment; filename="output_combination.json"' return response def kernel(ker, X1, X2, gamma): K = None if ker == 'linear': if X2 is not None: K = sklearn.metrics.pairwise.linear_kernel( np.asarray(X1), np.asarray(X2)) else: K = sklearn.metrics.pairwise.linear_kernel(np.asarray(X1)) elif ker == 'rbf': if X2 is not None: K = sklearn.metrics.pairwise.rbf_kernel( np.asarray(X1), np.asarray(X2), gamma) else: K = sklearn.metrics.pairwise.rbf_kernel( np.asarray(X1), None, gamma) return K class KMM: def __init__(self, kernel_type='linear', gamma=1.0, B=1.0, eps=None): ''' Initialization function :param kernel_type: 'linear' | 'rbf' :param gamma: kernel bandwidth for rbf kernel :param B: bound for beta :param eps: bound for sigma_beta ''' self.kernel_type = kernel_type self.gamma = gamma self.B = B self.eps = eps def fit(self, Xs, Xt): ''' Fit source and target using KMM (compute the coefficients) :param Xs: ns * dim :param Xt: nt * dim :return: Coefficients (Pt / Ps) value vector (Beta in the paper) ''' ns = Xs.shape[0] nt = Xt.shape[0] if self.eps == None: self.eps = self.B / np.sqrt(ns) K = kernel(self.kernel_type, Xs, None, self.gamma) kappa = np.sum(kernel(self.kernel_type, Xs, Xt, self.gamma) * float(ns) / float(n