ENH: provide a null_counts keyword to df.info() to force showing of null counts

[jreback]

Allows run-time control of showing of null-counts with df.info() (similar to what we allow with for example memory_usage in that this will override the default options for that call)

In [2]: df = DataFrame(1,columns=range(10),index=range(10))

In [3]: df.iloc[1,1] = np.nan

Default for a small frame (currently) is to show the non-null counts

In [5]: df.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 10 entries, 0 to 9
Data columns (total 10 columns):
0    10 non-null float64
1    9 non-null float64
2    10 non-null float64
3    10 non-null float64
4    10 non-null float64
5    10 non-null float64
6    10 non-null float64
7    10 non-null float64
8    10 non-null float64
9    10 non-null float64
dtypes: float64(10)
memory usage: 880.0 bytes

# allow it to be turned off
In [6]: df.info(null_counts=False)
<class 'pandas.core.frame.DataFrame'>
Int64Index: 10 entries, 0 to 9
Data columns (total 10 columns):
0    float64
1    float64
2    float64
3    float64
4    float64
5    float64
6    float64
7    float64
8    float64
9    float64
dtypes: float64(10)
memory usage: 880.0 bytes

When you have a big frame, right now you need to set max_info_rows to control this option

In [7]: pd.set_option('max_info_rows',5)           

In [8]: df.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 10 entries, 0 to 9
Data columns (total 10 columns):
0    float64
1    float64
2    float64
3    float64
4    float64
5    float64
6    float64
7    float64
8    float64
9    float64
dtypes: float64(10)
memory usage: 880.0 bytes

# force it to show
In [9]: df.info(null_counts=True)
<class 'pandas.core.frame.DataFrame'>
Int64Index: 10 entries, 0 to 9
Data columns (total 10 columns):
0    10 non-null float64
1    9 non-null float64
2    10 non-null float64
3    10 non-null float64
4    10 non-null float64
5    10 non-null float64
6    10 non-null float64
7    10 non-null float64
8    10 non-null float64
9    10 non-null float64
dtypes: float64(10)
memory usage: 880.0 bytes

[jreback]

@jorisvandenbossche I think this is a bug actually, e.g. if you have a really small info display it doesn't display any column counts, but prob a minor point.

[jorisvandenbossche]

I don't follow I think. Shouldn't this new kwarg null_counts not overrule if the frame is larger than max_info_rows/columns ?

[jreback]

it should and does, but the display is simply not printed out (too many columns), so this is sort of not-applicable.

[jorisvandenbossche]

ah, yes, I see, in such a case, the null_count option has just no effect. I don't really think it is a bug. You could warn that it is ignored in such case if it is specified, but doesn't seem that necesarilly.

[jreback]

yep, ok on the name null_counts? This is really just for an interactive session where you don't want to keep resetting max_info_rows.

[jorisvandenbossche]

yep, ok on the name!