Selecting a Single Column¶
Watch it
See the accompanied youtube video at the link here.
Something we often do in data analysis is obtain a single column from
a dataframe. We can again use .loc[] to do this which would look
something like this in general:
dataframe.loc[:, ['column name']]
So if we here want the column named type from our cereal dataframe we
could use the syntax:
cereal.loc[:, ['type']]
| type | |
|---|---|
| 0 | Cold |
| 1 | Cold |
| 2 | Cold |
| 3 | Cold |
| 4 | Cold |
| ... | ... |
| 72 | Cold |
| 73 | Cold |
| 74 | Cold |
| 75 | Cold |
| 76 | Cold |
77 rows × 1 columns
This seems a bit long winded and since we do this type of thing often. Luckily, Pandas has provided a quicker syntax to use to do the same thing.
Instead, selecting a single column can be done without using .loc[]
and we can just specify the dataframe name, followed by double square
brackets containing the column of interest (df[['column name']]).
cereal[['type']]
| type | |
|---|---|
| 0 | Cold |
| 1 | Cold |
| 2 | Cold |
| 3 | Cold |
| 4 | Cold |
| ... | ... |
| 72 | Cold |
| 73 | Cold |
| 74 | Cold |
| 75 | Cold |
| 76 | Cold |
77 rows × 1 columns
This makes the syntax for selecting the column type from the cereal
dataframe very easy.