python - 缓冲区 dtype 不匹配，预期为 'SIZE_t' 但得到 'long long'

Question

我在 spyder 中开发了 3 个 ML 模型，它们是线性回归、多项式回归和随机森林回归。在 sypder 中，它们都运行良好。但是，当我在 Django 上部署以创建 Web 应用程序时，随机森林引发了“ValueError: Buffer type mismatch, expected 'SIZE_t' but got 'long long'”。 (我尝试删除 randomforest，其他两个模型运行良好)。

先检查一下:-

在 Sypder 中开发的模型

"""****************** Import Lib ******************"""
import numpy as np
import pandas as pd
import seaborn as sns
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score

"""****************** Loading dataset ******************"""
boston = load_boston()
dataset = pd.DataFrame(boston.data, columns=boston.feature_names)
dataset['target'] = boston.target

"""****************** Data Preprocessing ******************"""
""" Data Analysis """
# Check Null
dataset.isnull().sum()
# Calculate X and y 
X = dataset.iloc[:,:-1].values
y = dataset.iloc[:,-1].values.reshape(-1,1)
# train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=25)
""" Visualizing Data """
corr = dataset.corr()
sns.heatmap(corr, annot=True, cmap='Blues')
sns.pairplot(dataset)

"""****************** Regression Models ******************"""
""" Linear Regression """
from sklearn.linear_model import LinearRegression
regressor_linear = LinearRegression()
regressor_linear.fit(X_train, y_train)
cv_linear = cross_val_score(estimator = regressor_linear, X=X_train, y=y_train, cv=10)

""" Polynomial Regression """
from sklearn.preprocessing import PolynomialFeatures
poly_reg = PolynomialFeatures(degree = 2)
X_poly = poly_reg.fit_transform(X_train)
regressor_poly2 = LinearRegression()
regressor_poly2.fit(X_poly, y_train)
cv_poly2 = cross_val_score(estimator=regressor_poly2, X=X_poly, y=y_train, cv=10)

""" Random Forest Regression """
from sklearn.ensemble import RandomForestRegressor
regressor_rf = RandomForestRegressor(n_estimators=500, random_state=0, n_jobs=-1)
regressor_rf.fit(X_train, y_train.ravel())
cv_rf = cross_val_score(estimator=regressor_rf, X=X_train, y=y_train.ravel(), cv=10)

"""****************** Measuring the Error ******************"""
models=[
        ('Linear Regression', cv_linear.mean()),
        ('Polynomial Regression (2)', cv_poly2.mean()),
        ('Random Forest Regression', cv_rf.mean())
        ]
cv_scores = pd.DataFrame(data=models, columns=['Model','CV Score'])

"""****************** Dump ******************"""
from sklearn.externals import joblib
joblib.dump(regressor_linear,'regressor_linear_jb')
joblib.dump(regressor_poly2,'regressor_poly2_jb')
joblib.dump(regressor_rf,'regressor_rf_jb')

Django 实现代码

from django.shortcuts import render
from django.http import HttpResponse
import json
from django.http import JsonResponse
import pandas as pd
import numpy as np
from sklearn.externals import joblib
from sklearn.preprocessing import PolynomialFeatures
# Create your views here.

# ML Code

regressor_linear = joblib.load('./models/regressor_linear_jb')
regressor_poly2 = joblib.load('./models/regressor_poly2_jb')
regressor_rf = joblib.load('./models/regressor_rf_jb')

# ML Code End

def predict(request):
    temp_data = [
                 0.16902,
                 0,
                 25.65,
                 0,
                 0.581,
                 5.986,
                 88.4,
                 1.9929,
                 2,
                 188,
                 19.1,
                 385.02,
                 14.81,
                 ]
    temp_df = pd.DataFrame(temp_data).transpose()
    predict = {}

    # Linear Regression
    predict['Linear Regressor'] = round(regressor_linear.predict(temp_df)[0, 0], 2)

    # Polynomial Regression.
    regressor_poly = PolynomialFeatures(degree=2)
    temp_df_poly = regressor_poly.fit_transform(temp_df)
    predict['Polynomial Regressor'] = round(regressor_poly2.predict(temp_df_poly)[0, 0], 2)
    
    # Random Forest Regression
    predict['Random Forest Regressor'] = round(regressor_rf.predict(temp_df)[0],2)
    
    return JsonResponse(predict)

Answer 1

将Django环境切换到anaconda即可解决

Jupyter notebook 使用的是 anaconda 环境，而 Django 使用的是安装在系统上的不同环境(主要问题 --> 一个是 32 位，而另一个是 64 位)