创建一个加载模型、获取数据向量并进行预测的函数

Question

我正在尝试在 Boston Housing 数据集上训练回归模型，并将模型保存到磁盘，然后创建 Title 函数。

下面的代码正在工作

import pandas as pd 
import numpy as np
import os
import pickle
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error
from joblib import dump, load 

boston = load_boston()
boston_df = pd.DataFrame(boston.data, columns = boston.feature_names)
y = pd.DataFrame(boston.target)
x_train, x_test, y_train, y_test = train_test_split(boston_df, y, test_size = 0.2, random_state=17)
model = LinearRegression()
model.fit(x_train, y_train)
y_predict = model.predict(x_test)
mse = mean_squared_error(y_test, y_predict)

with open('model.pkl', 'wb') as file:
    pickle.dump(model, file)
del model
with open('model.pkl', 'rb') as file:
    lin_model = pickle.load(file)

但是当我添加它时它应该工作

def predicting(x):
    x = pd.DataFrame(x)
    with open('model.pkl', 'rb') as file:
        lin_model = pickle.load(file)
        pred = lin_model.predict(x)
        print(pred)

但我不知道“x”需要什么结构才能使这个功能工作！

x = [1,2,4,5,6,3,5,6,7,3,5,6,4]
predicting(x)

像上面一样，它说 ValueError: matmul: 输入操作数 1 在其核心维度 0 中不匹配，具有 gufunc 签名 (n?,k),(k,m?)->(n?,m?) (尺寸 13 与 1) 不同

请帮忙？

Answer 1

import pandas as pd 
import numpy as np
import os
import pickle
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error
from joblib import dump, load 
from sklearn.datasets import make_blobs

boston = load_boston()
boston_df = pd.DataFrame(boston.data, columns = boston.feature_names)
y = pd.DataFrame(boston.target)
x_train, x_test, y_train, y_test = train_test_split(boston_df, y, test_size = 0.2, random_state=17)
model = LinearRegression()
model.fit(x_train, y_train)
y_predict = model.predict(x_test)
mse = mean_squared_error(y_test, y_predict)

with open('model.pkl', 'wb') as file:
    pickle.dump(model, file)
del model

x, _ = make_blobs(n_samples=3, centers=2, n_features=13, random_state=1)
def preficting(x):
    with open('model.pkl', 'rb') as file:
        lin_model = pickle.load(file)
        ynew = lin_model.predict(x)
        for i in range(len(x)):
            print("X=%s, Predicted=%s" % (x[i], ynew[i]))
preficting(x)

没有最终的印刷品，但我看到了它并认为它很漂亮，但这就是我需要的！ :)