These are present in the include/ directory and the implementations are present in the src/ directory.
C++ version of NumPy. Useful for building neural networks and statistical analyses. numcpp.h is the header file.
Includes several math functions in C++. math.h is the header file.
Preliminary version of a C++ deep learning library. cppdl.h is the header file. Work under progress.
Basic ML library. mllib.h is the header file. Work under progress.
A dynamically linked library (DLL) named libml.so is present in the lib/ directory. The definitions of all the relevant functions are present in this library and are loaded by the executable at runtime.
The g++ flags used to create the executable are
-O3 -Wall -Wextra -Wpedantic -I/path/to/include -L/path/to/lib -lml -fopenmp -Wl,-rpath=/path/to/lib
The -O3, -Wall, -Wextra, -Wpedantic flags are optional.
While building the library, if you have done sudo make install, all the flags except -fopenmp and -lml become optional.
You can also use make main=your_file.cpp target to build the executable.
To make the library again, the command is:
make Lib
To install it in the standard library and include directories, you can optionally enter sudo make install after doing this.
Special Thanks to Ashwin Abraham for refactoring the code and making the DLL (also for fixing random bugs).
Guys I really want to build this library much more and right now it's still in a very primary stage. If you happen to look at this repository, please please please contribute whatever you like. Even one PR would mean a lot to me. Let's make this world nicer by creating more and more open source stuff.