Dataframe

C++17 implementation of pandas.DataFrame

---

📖 Overview

A lightweight single-header C++17 dataframe implementation inspired by pandas.DataFrame.

Think of Dataframe as a spreadsheet in C++:

• Columns are named labels
• Rows are indexed records

Each column can hold different types (int, float, double, string), so you can load CSVs, manipulate data, and perform analytics with just a few lines of code — bringing pandas-like power to C++.

---

🛠 Requirements

• C++17 or newer
  Uses features like std::variant, std::optional, and std::from_chars.
• Windows console (only for colored logging via c_logger)

Note for Linux/Mac:
The dataframe core works on any platform.
c_logger uses Windows-specific APIs for colored output.
To run on Linux/Mac:

• Replace c_logger with a simple printf/std::cout logger, or
• Use ANSI escape codes for colors, or
• Stub out logging completely.

---

🚀 Installation

1. Add the single header to your project:

   dataframe.h
   

2. Include it in your code:

   #include "dataframe.h"

3. Compile with a C++17 compatible compiler.

---

Bash Build Instructions

MSVC (Windows Developer Command Prompt)

cl /std:c++17 /O2 main.cpp

MinGW (Windows)

g++ -std=gnu++17 -O2 main.cpp -o demo.exe

Linux / Mac (portable core; replace c_logger)

g++ -std=c++17 -O2 main.cpp -o demo

---

📂 Features

• Mixed types per column: using value_t = std::variant<double, std::string>;
• CSV I/O: from_csv(path, header) and to_csv(path, header)
• Column operations:
  • Add: add_column(name, values)
  • Rename: rename_column(old_name, new_name)
  • Remove: remove_column(name)
  • Shape: shape() -> {rows, columns}
• Row operations:
  • drop(index)
  • dropf() – drop first row
  • dropb() – drop last row
• Cleaning:
  • dropna() – remove rows with NaN
  • dropinf() – remove rows with Inf
  • dropemp() – remove rows with empty string cells
• Statistics:
  • sum, prod, mean, var, std, min, max
• Relationships:
  • cov – covariance
  • corr – correlation
• Higher moments:
  • skew – skewness
  • kurt – excess kurtosis
• Transforms:
  • diff – first difference
  • pct_change – percentage change
  • log_change – log returns
  • cumsum – cumulative sum
  • cumprod – cumulative product
• Selection:
  • at("col") -> std::vector<value_t>& (mutable reference to column)
  • at<T>("col") -> std::vector<T> typed extraction with automatic parsing
  • at<T>({col1, col2, ...}) -> std::vector<std::vector<T>> multiple columns
• Display:
  • print(n) – pretty console output
  • head(n) – first n rows
  • tail(n) – last n rows

---

📊 Example Usage

#include "dataframe.h"
#include <iostream>

int main() {
    c_dataframe df;

    // Add some columns
    df.add_column("Name", std::vector<std::string>{"Alice", "Bob", "Charlie"});
    df.add_column("Age", std::vector<int>{25, 30, 28});
    df.add_column("Salary", std::vector<double>{50000, 60000, 55000});

    // Print the dataframe
    df.print();

    // Access statistics
    double avg_salary = df.mean("Salary");
    double salary_std = df.std("Salary");
    std::cout << "Average salary: " << avg_salary << std::endl;
    std::cout << "Salary std dev: " << salary_std << std::endl;

    // Save to CSV
    df.to_csv("employees.csv", true);

    // Load from CSV
    c_dataframe df2("employees.csv", true);
    df2.print();

    return 0;
}

Example Console Output:

        Name                       Age                        Salary
-----------------------------------------------------------------------
0       Alice                      25                         50000
1       Bob                        30                         60000
2       Charlie                    28                         55000

📝 License

Licensed under the MIT License.