Dataframe Dimensions with Comma- Thousand Separator

[wamalab]

Please put a comma-thousand separator for a dataframe's dimensions. It will make the dimensions easier to read. In the attached image, "51,955x20 DataFrame" would be easier on the eye than the current format of "51955x20 DataFrame".

Thanks for all your good work on this package.

[MagicMuscleMan]

While I like the idea, please do not use a comma as a thousand separator. It is ambiguous and should not be used in an international context (which programming always is). A narrow no-break space would be the correct choice.

If you want to know more about the reasoning, you can use this Wikipedia section as an entry point to see how many standards advice against the comma.

In this specific case, I think it would be good to align this with the general formatting of Array dimensions. I would not see why the DataFrame dimensions should be handled differently to other Array dimensions.

[wamalab]

@MagicMuscleMan @bkamins We can always set the "default" by the user's region. I am in North America (and work in industry as a forensic accountant) and require a comma as a thousand separator. Perhaps Julia Dataframes can pick up a computer's settings to decide which thousand separator to use?

[bkamins]

@wamalab Maybe the easiest thing for you is to overwrite the default size display function:

Base.summary(df::AbstractDataFrame) =
    @sprintf("%d×%d %s", size(df)..., nameof(typeof(df)))

with whatever you find useful for your case. You can add this extra new function in your scripts after you load DataFrames.jl.