m2cgen 简介

m2cgen（Model 2 Code Generator）-是一个轻量级库，它提供了一种将经过训练的统计模型转换为本机代码（Python、C、Java、Go、JavaScript、Visual Basic、C#、PowerShell、R、PHP、Dart、Haskell、Ruby、F#、Rust）的简便方法。

简而言之，它可以将python scikit-learn 等训练的机器学习模型转成C，JAVA等能够直接运行的代码，从而在无需依赖库的情况下直接运行

m2cgen 安装

注意python 版本需要在3.6以上

pip install m2cgen

示例代码

from sklearn.datasets import load_iris

from xgboost.sklearn import XGBClassifier
from xgboost import plot_importance

import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

import m2cgen as m2c

# 记载样本数据集
iris = load_iris()
x,y = iris.data,iris.target

# 数据集分割
x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.2,random_state=123457)

clf = XGBClassifier(
    booster = 'gbtree',
    objective = 'multi:softmax',
    num_class = 3,
    gamma = 0.1,
    max_depth = 6,
    reg_lambda = 2,
    subsample = 0.7,
    colsample_bytree = 0.7,
    min_child_weight = 3,
    eta = 0.1,
    seed = 1000,
    nthread = 4,
)

#训练模型
clf.fit(x_train,y_train,eval_metric='auc')

#模型转为c代码
code = m2c.export_to_c(clf)
print(code)

转化结果是一段很长的c代码，这里只粘贴部分内容

void score(double * input, double * output) {
    double var0;
    if ((input[f2]) >= (2.6)) {
        var0 = -0.06913044;
    } else {
        var0 = 0.13098592;
    }
    double var1;
    if ((input[f3]) >= (0.8)) {
        var1 = -0.06719698;
    } else {
        var1 = 0.11539519;
    }
    double var2;
    if ((input[f3]) >= (0.8)) {
        var2 = -0.06486788;
    } else {
        var2 = 0.104500055;
    }