mltemplate

🚀 Modern C++17 Machine Learning Template

A complete starter template for building machine learning projects in pure C++17.
Combines mlcore, dataframe, and matplotlib-cpp into a single ready-to-run example.

---

📖 Overview

mltemplate is a modern C++17 machine learning template that integrates:

• mlcore → ML algorithms (Random Forest, SVM, KNN, etc.)
• dataframe → Pandas-like DataFrame in C++
• matplotlib-cpp → C++ wrapper for Python’s Matplotlib

It demonstrates an end-to-end ML pipeline in C++: load a dataset, preprocess features, train a model, visualize results, and export predictions to CSV. The repository includes a working Random Forest example on the Pima Indians Diabetes dataset.

---

🛠 Requirements

• C++17 or newer
• Python with Matplotlib (required by matplotlib-cpp)

Dependencies

• mlcore
• dataframe
• matplotlib-cpp

---

🚀 Installation

# Clone repository
git clone --recursive https://github.com/overvac/mltemplate.git
cd mltemplate

# Build and run (MSVC / GCC / Clang)
# Windows (MSVC)
cl /std:c++17 /O2 main.cpp

# Linux / macOS
g++ -std=c++17 -O2 main.cpp -o app
./app

▶️ Example Program

#include "mltemplate/mltemplate.h"

int main()
{
    c_dataframe df("diabetes.csv");
    df.print();

    std::vector<std::string> feature_column_names = {
        "Pregnancies", "Glucose", "BloodPressure", "SkinThickness",
        "Insulin", "BMI", "DiabetesPedigreeFunction", "Age"
    };

    auto features = n_mlcore::n_preprocessing::c_transform::get().transpose<double>(
        df.at<double>(feature_column_names)
    );
    auto outcomes = df.at<int>("Outcome");

    n_mlcore::n_supervised::c_rf forest(500);
    forest.fit(features, outcomes);

    auto fi = forest.feature_importances();
    n_plt::bar(fi);
    n_plt::show();

    std::vector<int> predicted;
    predicted.reserve(features.size());
    for (const auto& f : features)
    {
        int p = forest.predict({ f })[0];
        predicted.push_back(p);
    }

    df.add_column<int>("Predicted", predicted);
    df.to_csv("solved_diabetes.csv");

    return 0;
}

📊 Output

• Prints dataset summary from diabetes.csv
• Trains a Random Forest classifier (mlcore)
• Plots feature importances (matplotlib-cpp)
• Saves predictions to solved_diabetes.csv

📜 License

Released under the MIT License.