Fast Linear Regression in C

Clone and install

git clone https://github.com/alessandrobessi/fast-linear-regression.git
cd fast-linear-regression
cd src
make
make clean

Usage:

1. Generate sample data. The following command produces two csv files:

• X_train.csv: a matrix of dimensions (num_examples, num_features) with training examples
• y_train.csv: a matrix of dimensions (num_examples, 1) with training labels

./fast-lr generate [num_features] [num_examples]

2. Fit a linear regression model. The following command estimates the linear regression coefficients starting from two csv file, one containing the training examples and the other one containing the training labels.

./fast-lr fit [X_train_csv_file] [y_train_csv_file] [--verbose] [--with-intercept]

3. Predict. The following command predicts y values using two csv file, one containing training examples and the other one containing linear regression coefficients.

./fast-lr predict [X_train_csv_file] [beta_csv_file] [--verbose] [--with-intercept]

Example:

./fast-lr generate 2000 100000
./fast-lr fit X_train.csv y_train.csv --verbose
./fast-lr predict X_train.csv beta.csv --verbose

./fast-lr fit X_train.csv y_train.csv --verbose --with-intercept
./fast-lr predict X_train.csv beta.csv --verbose --with-intercept