Python新手求教：怎样把两个列向量合并成一个 n*2 的矩阵？
最近学习随即森林分类算法，碰到一个问题，试了各种互联网上的方法，都不能得到正确结果，只好在这里求助大家了.
是这样：test_lables 是测试样本二分类的真实标签，有 692 个样本，test_hat 是预测值，现在我想把这两个合并在一块，组成一个 6922 的矩阵，每个预测值对应一个真实值。源代码如下：
import numpy as np
import pandas as pd
from sklearn.ensemble import RandomForestClassifier
from sklearn.dummy import DummyClassifier
from sklearn.ensemble import AdaBoostClassifier
from sklearn.svm import SVC
#from sklearn import datasets
dataframe = pd.read_csv( “D:/Research/TuPo_sel0.Train.csv”, header = None )
train_features = dataframe.iloc[ :, 0:24]
train_lables = dataframe.iloc[:, 24]
test_data = pd.read_csv( “D:/Research/TuPo_sel0.Valid.csv”, header = None )
test_features = test_data.iloc[ :, 0:24 ]
test_lables = test_data.iloc[ :, 24 ]
dummy = DummyClassifier( strategy = ‘uniform’, random_state = 1 )
dummy.fit( train_features, train_lables )
print( “dummy_score =”, dummy.score( test_features, test_lables ) )
style = 1
if style == 1:
max_features = 19
n_estimators = 400
randomforest = RandomForestClassifier( max_features = max_features, n_estimators = n_estimators, random_state=1, n_jobs=-1 )
model = randomforest.fit( train_features, train_lables )
test_hat = model.predict( test_features )
test_hat1 = np.hstack( ( test_hat, test_lables ) )
test_hat1.reshape( -1, 2 )
print( test_hat1.shape )
print( test_hat1 )
print( “max_features =”, max_features, “; n_estimators =”, n_estimators,
“; randomforest_score =”, randomforest.score( test_features, test_lables ) )
运算结果如下：
runfile(‘D:/Python Programs/TryLoadData.py’, wdir=‘D:/Python Programs’)
dummy_score = 0.5447976878612717
(1384,)
[0 0 1 … 0 0 0]
max_features = 19 ; n_estimators = 400 ; randomforest_score = 0.6416184971098265
求教各位怎么修改才能得到正确结果？
Python新手求教：怎样把两个列向量合并成一个 n2 的矩阵？
import numpy as np
假设你有两个列向量
col1 = np.array([1, 2, 3]) col2 = np.array([4, 5, 6])
方法1：使用np.column_stack() - 最直观
matrix1 = np.column_stack((col1, col2))
方法2：使用np.hstack() - 水平堆叠
matrix2 = np.hstack((col1.reshape(-1, 1), col2.reshape(-1, 1)))
方法3：使用np.concatenate()
matrix3 = np.concatenate((col1.reshape(-1, 1), col2.reshape(-1, 1)), axis=1)
方法4：先转置再转置（如果向量是行向量形式）
假设向量是行向量
row1 = np.array([[1, 2, 3]]) row2 = np.array([[4, 5, 6]]) matrix4 = np.vstack((row1, row2)).T
print(“方法1结果：”) print(matrix1) print("\n形状：", matrix1.shape)
验证所有方法结果相同
print("\n所有方法结果是否一致：", np.array_equal(matrix1, matrix2) and np.array_equal(matrix2, matrix3) and np.array_equal(matrix3, matrix4))
推荐用np.column_stack()，最符合直觉。
test_hat1 = np.hstack((test_hat.reshape(-1, 1), test_lables.reshape(-1, 1)))
查看训练结果可以看简报，metrics.classification_report
