Skip to content

Latest commit

 

History

History
60 lines (43 loc) · 3.36 KB

README.md

File metadata and controls

60 lines (43 loc) · 3.36 KB

arrgh

A small python utility to pretty-print a table summarizing arrays/tensors/scalars from numpy, pytorch, etc.

Why the name? "arr" is short for "array", and "arrgh" is the sound you make while debugging your array shapes.

Example

Calling arrgh(my_arr1, my_arr2, ...) prints a table like:

       name | dtype         | shape         | type          | device | min         | max         | mean       
--------------------------------------------------------------------------------------------------------------
     [None] | None          | N/A           | NoneType      |        | N/A         | N/A         | N/A        
    intval1 | int           | scalar        | int           |        |      7      |      7      |      7     
    intval2 | int           | scalar        | int           |        |     -3      |     -3      |     -3     
  floatval0 | float         | scalar        | float         |        |     42      |     42      |     42     
  floatval1 | float         | scalar        | float         |        | 5.5e-12     | 5.5e-12     | 5.5e-12    
  floatval2 | float         | scalar        | float         |        | 7.72324e+44 | 7.72324e+44 | 7.72324e+44
     npval1 | int64         | [100]         | numpy.ndarray |        |      0      |     99      |   49.5     
     npval2 | int64         | [10000]       | numpy.ndarray |        |      0      |   9999      | 4999.5     
     npval3 | uint64        | [10000]       | numpy.ndarray |        |      0      |   9999      | 4999.5     
     npval4 | float32       | [100, 10, 10] | numpy.ndarray |        |      0      |   9999      | 4999.5     
[temporary] | float32       | [10, 8]       | numpy.ndarray |        |      2      |     99      |   50.5     
     npval5 | int64         | []            | numpy.int64   |        |   9999      |   9999      |   9999     
  torchval1 | torch.float32 | [1000, 12, 3] | torch.Tensor  | cpu    | -4.08445    | 3.90982     | 0.00404567 
  torchval2 | torch.float32 | [1000, 12, 3] | torch.Tensor  | cuda:0 | -3.87309    | 3.90342     | 0.00339224 
  torchval3 | torch.int64   | [1000]        | torch.Tensor  | cpu    |      0      |    999      | N/A        
  torchval4 | torch.int64   | []            | torch.Tensor  | cpu    |      0      |      0      | N/A

Use keyword arguments like arrgh(custom_name=3.*my_arr1) to specify names explicitly or give names to temporaries.

Installation

pip install arrgh

from arrgh import arrgh

Docs

The package exposes a single function called arrgh(). Call it like: arrgh(my_arr, some_other_arr, maybe_a_scalar).

The function accepts a variable number of arguments. Arrays can also be passed as named optional arguments.

Inputs can be:

  • Numpy tensor arrays
  • Pytorch tensor arrays
  • Jax tensor arrays
  • Python ints / floats
  • None
  • It may also work with other array-like types, but they have not been tested.
  • Input values which are not arrays or numeric types (strings, objects, etc) will be printed as blank rows in the table.

When arrays are passed as variable arguments, the printed name of the array in the table is inferred from the variable name in the outer scope, when possible. When arrays are passed as named keyword arguments, the key name is used.

Pass an integer for the arrgh_float_width option to specify the precision to which floating point types are printed.

Author: Nicholas Sharp (nmwsharp.com)