基于逻辑回归的心脏病预测

from __future__ import print_function
from sklearn.metrics import confusion_matrix
from sklearn.linear_model import LogisticRegressionCV
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model.coordinate_descent import ConvergenceWarning
import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt
from sklearn.externals import joblib
from sklearn.metrics import accuracy_score
from sklearn.metrics import roc_curve
from sklearn.metrics import auc
import seaborn as sns
import warnings

# 不显示警告信息
warnings.filterwarnings(action='ignore', category=ConvergenceWarning)
# plt中显示中文的问题
mpl.rcParams["font.sans-serif"] = [u"SimHei"]
mpl.rcParams["axes.unicode_minus"] = False

# 读取文件
data = pd.read_csv('./heart.csv')

# 分出特征和标签
X = data.iloc[:, :-1]
y = data.iloc[:, -1].to_frame()
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.25, random_state=42)
# X_train = (X_train - np.min(X_train)) / (np.max(X_train) - np.min(X_train)).values
# X_test = (X_test - np.min(X_test)) / (np.max(X_test) - np.min(X_test)).values

# 对数据的训练集进行标准化
ss = StandardScaler()
X_train = ss.fit_transform(X_train)

# ===================================================
# 分析特征部分，主要为画图
# ===================================================
sns.pairplot(data, hue='target')
plt.show()

g = sns.jointplot(data.thalach, data.chol, kind="kde", height=7, space=0)
plt.show()
# ===================================================
sns.swarmplot(y="cp", x="restecg", hue="thalach",
              palette=["r", "c", "y"], data=data)
plt.show()
# ====================================================
fig = sns.regplot(x="oldpeak", y="thalach", data=data)
plt.title("GRE Score vs CGPA")
plt.show()
# ===================================================
data.describe().plot(kind="area", fontsize=27, figsize=(20, 8), table=True, colormap="rainbow")
plt.xlabel('Statistics', )
plt.ylabel('Value')
plt.title("General Statistics of Admissions")
plt.show()

sns.lineplot(x="age", y="trestbps",
             data=data, color='b', label='trestbps')
sns.lineplot(x="age", y="chol",
             data=data, color='r', label='chol')
sns.lineplot(x="age", y="thalach",
             data=data, color='G', label='thalach')
plt.legend(loc=2)
plt.show()
# =========================
plt.figure(figsize=(20,25))
i = 0

for item in data.columns:
    i += 1
    plt.subplot(5, 3, i)
    sns.distplot(data[item], rug=True, rug_kws={"color": "b"},kde=True,
                 kde_kws={"color": "blue", "lw": 5, "label": "KDE"},
                 hist_kws={"histtype": "step", "linewidth": 3,"alpha": 1, "color": "blue"},label="{0}".format(item))
#     sns.distplot(admission_v1[item], kde=True,label="{0}".format(item))
plt.show()
# ==============================
for i,col in enumerate(data.columns):
    plt.subplot(5,3,i+1)
    plt.scatter(np.arange(1,304),data[col].values.tolist())
    plt.title(col)
    fig, ax = plt.gcf(), plt.gca()
    fig.set_size_inches(10, 10)
    plt.tight_layout()
plt.show()
# ==============================

# 逻辑回归的参数设定
parameters = [
    {
        'penalty': ['l1', 'l2'],
        'C': [0.1, 0.4, 0.5],
        'random_state': [0]
    },
]


# lr = LogisticRegression(C=0.1, penalty='l1', random_state=0)
# lr.fit(X_train, y_train)

# ===========================================================================
# 构建并训练模型
# ===========================================================================
def plot_roc_(false_positive_rate, true_positive_rate, roc_auc):
    """
    :param false_positive_rate:
    :param true_positive_rate:
    :param roc_auc:
    :return: 画roc曲线
    """
    plt.figure(figsize=(5, 5))
    plt.title('Receiver Operating Characteristic')
    plt.plot(false_positive_rate, true_positive_rate, color='red', label='AUC = %0.2f' % roc_auc)
    plt.legend(loc='lower right')
    plt.plot([0, 1], [0, 1], linestyle='--')
    plt.axis('tight')
    plt.ylabel('True Positive Rate')
    plt.xlabel('False Positive Rate')
    plt.show()


def plot_feature_importances(gbm):
    """
    :param gbm:
    :return: 画出特征图
    """
    n_features = X_train.shape[1]
    plt.barh(range(n_features), gbm.feature_importances_, align='center')
    plt.yticks(np.arange(n_features), X_train.columns)
    plt.xlabel("Feature importance")
    plt.ylabel("Feature")
    plt.ylim(-1, n_features)


lr = LogisticRegressionCV(multi_class="ovr", fit_intercept=True, Cs=np.logspace(-2, 2, 20), cv=2, penalty="l2",
                          solver="lbfgs", tol=0.01)
re = lr.fit(X_train, y_train)

# 模型效果获取

r = re.score(X_train, y_train)
print("R值(准确率):", r)
print("参数:", re.coef_)
print("截距:", re.intercept_)
print("稀疏化特征比率:%.2f%%" % (np.mean(lr.coef_.ravel() == 0) * 100))
print("=========sigmoid函数转化的值，即：概率p=========")
print(re.predict_proba(X_test))  # sigmoid函数转化的值，即：概率p

# 模型的保存与持久化

joblib.dump(ss, "logistic_ss.model")  # 将标准化模型保存
joblib.dump(lr, "logistic_lr.model")  # 将训练后的线性模型保存
joblib.load("logistic_ss.model")  # 加载模型,会保存该model文件
joblib.load("logistic_lr.model")

# 预测
X_test = ss.transform(X_test)  # 数据标准化
Y_predict = lr.predict(X_test)  # 预测

# 画图对预测值和实际值进行比较
x = range(len(X_test))
plt.figure(figsize=(14, 7), facecolor="w")
plt.ylim(-1, 2)
plt.plot(x, y_test, "ro", markersize=8, zorder=3, label=u"真实值")
plt.plot(x, Y_predict, "go", markersize=14, zorder=2, label=u"预测值,$R^2$=%.3f" % lr.score(X_test, y_test))
plt.legend(loc="upper left")
plt.xlabel(u"样本编号", fontsize=18)
plt.ylabel(u"是否患有心脏病", fontsize=18)
plt.title(u"Logistic算法对数据进行分类", fontsize=20)
plt.savefig("Logistic算法对数据进行分类.png")
plt.show()

print("=============Y_test==============")
print(np.array(y_test).ravel())
print("============Y_predict============")
print(Y_predict)

y_pred = re.predict(X_test)
y_proba = re.predict_proba(X_test)

false_positive_rate, true_positive_rate, thresholds = roc_curve(y_test, y_proba[:, 1])
roc_auc = auc(false_positive_rate, true_positive_rate)
plot_roc_(false_positive_rate, true_positive_rate, roc_auc)

# 查看模型性能
# print('Hata Oranı :',r2_score(y_test,y_pred))
print('Accurancy Oranı :', accuracy_score(y_test, y_pred))
print("Logistic TRAIN score with ", format(re.score(X_train, y_train)))
print("Logistic TEST score with ", format(re.score(X_test, y_test)))
print()

cm = confusion_matrix(y_test, y_pred)
print(cm)
sns.heatmap(cm, annot=True)
plt.show()

print('CoEf:\n')
print(re.coef_)
print('Intercept_\n')
print(re.intercept_)
print('Proba:\n')
print(re.predict_log_proba)