Smarter Debug formatting for arrays

[bluss]

• Print only part of the array if it has too many elements
• Select width and precision better, for a nice and readable display

inspiration:

array([[ 0.    ,  0.0001,  0.0002, ...,  0.0002,  0.0003,  0.0004],
       [ 0.1   ,  0.09  ,  0.1   , ...,  0.1   ,  0.1   ,  0.1   ],
       [ 0.0024,  0.0025,  0.0026, ...,  0.0026,  0.0027,  0.0028],
       ..., 
       [ 0.0024,  0.0025,  0.0026, ...,  0.0026,  0.0027,  0.0028],
       [ 0.0036,  0.0037,  0.0038, ...,  0.0038,  0.0039,  0.004 ],
       [ 0.0048,  0.0049,  0.005 , ...,  0.005 ,  0.0051,  0.0052]])

[andrei-papou]

Hi @bluss , @LukeMathWalker . I've never contributed to ndarray before and since it is a good first issue now I would like to pick this up. Could you please provide any guidance? Thanks.

[LukeMathWalker]

Welcome @andrei-papou!

The Debug trait defines how a Rust variable is printed when using the {:?} format placeholder (see here for a more complete overview).

You can find ArrayBase's implementation of the Debug trait here - the bulk of the job takes place in the format_array function.
As you can see, it prints all the array elements to screen, which turn out to be quite noisy if you are working with big arrays or you are printing arrays from more than one spot in your program.

You can get a taste of what this looks like using 1-dimensional arrays:

extern crate ndarray;

use ndarray::Array;

fn main() {
    let v = Array::linspace(0., 1., 128);
    println!("{:?}", v);
}

This produces:

[0.0, 0.0078125, 0.015625, 0.0234375, 0.03125, 0.0390625, 0.046875, 0.0546875, 0.0625, 0.0703125, 0.078125, 0.0859375, 0.09375, 0.1015625, 0.109375, 0.1171875, 0.125, 0.1328125, 0.140625, 0.1484375, 0.15625, 0.1640625, 0.171875, 0.1796875, 0.1875, 0.1953125, 0.203125, 0.2109375, 0.21875, 0.2265625, 0.234375, 0.2421875, 0.25, 0.2578125, 0.265625, 0.2734375, 0.28125, 0.2890625, 0.296875, 0.3046875, 0.3125, 0.3203125, 0.328125, 0.3359375, 0.34375, 0.3515625, 0.359375, 0.3671875, 0.375, 0.3828125, 0.390625, 0.3984375, 0.40625, 0.4140625, 0.421875, 0.4296875, 0.4375, 0.4453125, 0.453125, 0.4609375, 0.46875, 0.4765625, 0.484375, 0.4921875, 0.5, 0.5078125, 0.515625, 0.5234375, 0.53125, 0.5390625, 0.546875, 0.5546875, 0.5625, 0.5703125, 0.578125, 0.5859375, 0.59375, 0.6015625, 0.609375, 0.6171875, 0.625, 0.6328125, 0.640625, 0.6484375, 0.65625, 0.6640625, 0.671875, 0.6796875, 0.6875, 0.6953125, 0.703125, 0.7109375, 0.71875, 0.7265625, 0.734375, 0.7421875, 0.75, 0.7578125, 0.765625, 0.7734375, 0.78125, 0.7890625, 0.796875, 0.8046875, 0.8125, 0.8203125, 0.828125, 0.8359375, 0.84375, 0.8515625, 0.859375, 0.8671875, 0.875, 0.8828125, 0.890625, 0.8984375, 0.90625, 0.9140625, 0.921875, 0.9296875, 0.9375, 0.9453125, 0.953125, 0.9609375, 0.96875, 0.9765625, 0.984375, 0.9921875, 1.0] shape=[129], strides=[1], layout=C | F (0x3), const ndim=1

It would look much better if it were to be something like:

[0.0, 0.0078125, 0.015625, ..., 0.984375, 0.9921875, 1.0] 
shape=[129], strides=[1], layout=C | F (0x3), const ndim=1

When you move to multi-dimensional arrays additional issues arise (e.g. alignment of array elements), but I'd say that a good starting point could be to reduce the number of printed elements adding ellipsis (...) to signal that something has been omitted to reduce visual noise.

I am happy to provide you with additional info if you need!

[andrei-papou]

@LukeMathWalker Thanks a lot, everything is pretty clear! I'll submit a PR soon.

[dineshadepu]

Hi @LukeMathWalker. I am trying to contribute to ndarray. To get started I had a look at arrayformat file, which implements the printing. In that file I don't see where the format function is implemented. Is this in std lib? https://github.com/rust-ndarray/ndarray/blob/master/src/arrayformat.rs#L74.

[LukeMathWalker]

Yes @dineshadepu, format is a macro in Rust's standard library - you can find the relevant documentation here.
Given that @andrei-papou is already working on this issue, may I suggest you to have a look at #597 for other needed contributions to the library?

[andrei-papou]

@LukeMathWalker I've submitted a PR #606
Could you please take a look or assign someone as a reviewer? Thanks a lot.

[jturner314]

Fixed by #606.