Goopy

A simple, lightweight NumPy-like array library implemented in C for basic multi-dimensional array operations.

Features

• supports n-dimensional arrays
• broadcasting support for basic arithmetic operations
• batched matrix multiplication

#include <goopy.h>

int main() {

  // TESTCASE - tranpose()
  array_t a = arange(1, 7, 1);
  reshape(&a, (size_t[]){1, 2, 3}, 3);
  PRINT_ARRAY(a);

  array_t b = arange(1, 7, 1);
  reshape(&b, (size_t[]){3, 2, 1}, 3);
  PRINT_ARRAY(b);

  array_t c = element_wise_mul(&a, &b);
  PRINT_ARRAY(c);

  deinit_array(&a);
  deinit_array(&b);
  deinit_array(&c);
  return 0;
}

API Reference

Array Creation

• arange(start, stop, step) - Create 1D array with range of values
• init_array_with_zeros(shape, ndim) - Create array filled with zeros
• init_array_with_ones(shape, ndim) - Create array filled with ones
• init_array_with_scalar_value(shape, ndim, value) - Create array filled with scalar

Operations

• element_wise_add(a, b) - Element-wise addition with broadcasting
• element_wise_sub(a, b) - Element-wise subtraction with broadcasting
• element_wise_mul(a, b) - Element-wise multiplication with broadcasting
• element_wise_div(a, b) - Element-wise division with broadcasting
• matmul(a, b) - Matrix multiplication with batch support

Array Manipulation

• reshape(arr, new_shape, new_ndim) - Reshape array
• transpose(arr) - Transpose array (reverse all dimensions)
• flatten(arr) - Flatten to 1D array

Utilities

• PRINT_ARRAY(arr) - Print array contents
• deinit_array(arr) - Free array memory

Notes

• Arrays use int data type only
• Memory management is manual - always call deinit_array() for arrays you create
• Broadcasting follows NumPy-like rules
• Matrix multiplication works on the last two dimensions for batched operations