Two popular options are to use:
ipython notation:
In [11]: df = pd.DataFrame([[1, 2], [3, 4]])
In [12]: df
Out[12]:
0 1
0 1 2
1 3 4
Alternatively (this is popular over in the python documentation) and more concisely:
df.columns # Out: RangeIndex(start=0, stop=2, step=1)
df[0]
# Out:
# 0 1
# 1 3
# Name: 0, dtype: int64
for col in df:
print(col)
# prints:
# 0
# 1
Generally, this is better for smaller examples.
Note: The distinction between output and printing. ipython makes this clear (the prints occur before the output is returned):
In [21]: [print(col) for col in df]
0
1
Out[21]: [None, None]