Some simple examples.

This directory contain some examples of how to save and load Numpy arrays with this library.

Compile the four examples with make.

$ make

This will build four executables in the same directory.

how_to_save

Please study the code in how_to_save.c. If you run the code it will generate two .npy files.

• my_4_by_3_array.npy
• my_4_by_3_array_shortcut.npy

Please open a Python REPL, and make sure you can read these.

how_to_save_npz

Please study the code in how_to_save_npz.c. If you run the code it will generate a .npz file containing two NumPy arrays.

Please open a Python REPL, and make sure you can read the .npz file, and that the arrays looks sensible.

how_to_load

If you study the code in how_to_load.c, you will see how simple it is to load an array from a file. The example code even shows how you can inteact with the array data. To run the example it takes a filename as a command line argument. If you ran the above how_to_save-example, you can run:

$ ./how_to_load my_4_by_3_array.npy

This will load the array and dump some info, and write if the number format is float32.

You can also try to generate a NumPy file from your Python REPL, and load it with the example.

how_to_load_npz

Please study the code in how_to_load_npz.c. It loads a .npz file into memory and returns all arrays in a linked list.

You can run the example with the .npzfile generated my how_to_save_npz:

$ ./how_to_load_npz iarray_and_darray.npz

You can also create a .npz file from your Python REPL, and check that it can be read as well.

Loading with npy_array_list_load() allocate all memory for all holding list structures, all npy_array structures and all the numeric data itself. Later calling npy_array_list_free() with the head element of the list as argument, will free everything.