[已解决]求助，怎么实现在固定的数组中随机抽取几个数，重复这个过程几次，这几次不能重复

乌克大喵喵

一共有3种花，每种花50个数据，现在的参数是每种花前30个为训练组，后二十个为实验组
test1是30个里随机挑20个作为训练组然后制成knn，这个过程重复20次，但需要这20次的数据不能完全相同。
这个过程该如何实现


import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from scipy.spatial.distance import cdist


def SplitData_1(data):
    traindata, testdata = [], []
    for i in range(3):
        idx = np.arange(50)
        trainidx, testidx = idx[:30], idx[30:]
        trainidx=np.random.permutation(trainidx)
        datai = data.loc[data.iloc[:, -1] == i, :]
        traindata.append(datai.iloc[trainidx, :])
        testdata.append(datai.iloc[testidx, :])

    traindata = pd.concat(traindata, 0)
    testdata = pd.concat(testdata, 0)


    #print('traindata:', traindata )
    #print('testdata:', testdata )

    Xtrain = traindata.iloc[:, :-1]
    ytrain = traindata.iloc[:, -1]
    Xtest = testdata.iloc[:, :-1]
    ytest = testdata.iloc[:, -1]

    return Xtrain.values, Xtest.values, ytrain.values, ytest.values

def SplitData_2(data):
    traindata, testdata = [], []
    for i in range(3):
        idx = np.arange(50)
        trainidx, testidx = idx[:30], idx[30:]
        trainidx=np.random.permutation(trainidx)
        trainidx=trainidx[:20]
        datai = data.loc[data.iloc[:, -1] == i, :]
        traindata.append(datai.iloc[trainidx, :])
        testdata.append(datai.iloc[testidx, :])

    traindata = pd.concat(traindata, 0)
    testdata = pd.concat(testdata, 0)


    #print('traindata:', traindata )
    #print('testdata:', testdata )

    Xtrain = traindata.iloc[:, :-1]
    ytrain = traindata.iloc[:, -1]
    Xtest = testdata.iloc[:, :-1]
    ytest = testdata.iloc[:, -1]

    return Xtrain.values, Xtest.values, ytrain.values, ytest.values

def SplitData_3(data):
    traindata, testdata = [], []
    for i in range(3):
        idx = np.arange(50)
        trainidx, testidx = idx[:30], idx[30:]
        trainidx=np.random.permutation(trainidx)
        trainidx = trainidx[:25]
        datai = data.loc[data.iloc[:, -1] == i, :]
        traindata.append(datai.iloc[trainidx, :])
        testdata.append(datai.iloc[testidx, :])

    traindata = pd.concat(traindata, 0)
    testdata = pd.concat(testdata, 0)


    #print('traindata:', traindata )
    #print('testdata:', testdata )

    Xtrain = traindata.iloc[:, :-1]
    ytrain = traindata.iloc[:, -1]
    Xtest = testdata.iloc[:, :-1]
    ytest = testdata.iloc[:, -1]

    return Xtrain.values, Xtest.values, ytrain.values, ytest.values


class KNN(object):
    def __init__(self, k):
        self.k = k

    def fit(self, Xtrain, ytrain):
        self.Xtrain = Xtrain
        self.ytrain = ytrain

        return self

    def predict(self, Xtest):
        k = self.k
        Xtrain = self.Xtrain
        ytrain = self.ytrain
        N = Xtest.shape[0]
        y_pred = np.zeros([N, 1])
        for i in range(Xtest.shape[0]):
            x = Xtest[i, :]  # 提取预测样本
            # 计算样本训练集中各个样本的距离
            dists = cdist(Xtrain, np.expand_dims(x, 0))
            # 对距离进行升序排列，记录排序后的位置
            idxs = np.argsort(dists, 0)
            # 取排在前k个的训练集中样本对应的索引，提取他们的label,并用其中的众数作为预测结果
            cand = pd.DataFrame(data=ytrain[idxs[:k]])
            y_pred[i] = cand.mode()[0]

        return y_pred



# %% Load and split data
data = pd.read_csv('iris.data', header=None)
class_map = {'Iris-setosa': 0, 'Iris-versicolor': 1, 'Iris-virginica': 2}
data.iloc[:, -1] = data.iloc[:, -1].map(class_map)

# %% KNN
from sklearn.metrics import accuracy_score
acc_knn_1 = np.zeros(20)#实验次数
acc_knn_2 = np.zeros(20)
acc_knn_3 = np.zeros(20)
for i in range(20):
    Xtrain, Xtest, ytrain, ytest = SplitData_1(data)
    knn = KNN(k=3).fit(Xtrain, ytrain)
    ypred = knn.predict(Xtest)
    acc_knn_1[i] = accuracy_score(ypred, ytest)
print('KNN Result:')
print(acc_knn_1)
print('Acc mean:', np.mean(acc_knn_1))
print('Acc std:', np.std(acc_knn_1))


# %% Test1
for j in range(20):
    Xtrain, Xtest, ytrain, ytest = SplitData_2(data)
    knn = KNN(k=3).fit(Xtrain, ytrain)
    ypred = knn.predict(Xtest)
    acc_knn_2[j] = accuracy_score(ypred, ytest)
print('KNN Result test1:')
print(acc_knn_2)
print('Acc mean:', np.mean(acc_knn_2))
print('Acc std:', np.std(acc_knn_2))


# %% Test2
for k in range(20):
    Xtrain, Xtest, ytrain, ytest = SplitData_3(data)
    knn = KNN(k=3).fit(Xtrain, ytrain)
    ypred = knn.predict(Xtest)
    acc_knn_3[k] = accuracy_score(ypred, ytest)
print('KNN Result test2:')
print(acc_knn_3)
print('Acc mean:', np.mean(acc_knn_3))
print('Acc std:', np.std(acc_knn_3))

最佳答案

月排行榜 / 总排行榜

qq1151985918

2021-6-3 15:24:43

参考代码

from random import sample

data_rand_history = []
while len(data_rand_history) < 20:
    rand = set(sample(range(30), 20))
    if rand not in data_rand_history:
        data_rand_history.append(rand)
print(data_rand_history)

跳转到最佳答案楼层

qq1151985918

参考代码

from random import sample

data_rand_history = []
while len(data_rand_history) < 20:
    rand = set(sample(range(30), 20))
    if rand not in data_rand_history:
        data_rand_history.append(rand)
print(data_rand_history)

乌克大喵喵

在SplitData_2中加了以下条件
trainidx = sorted(trainidx)
    if trainidx not in data_rand_history2:
        data_rand_history2.append(trainidx)
    else:
        SplitData_2(data)