按标签选择列

# Create a sample DF
df = pd.DataFrame(np.random.randn(5, 3), columns=list('ABC'))

# Show DF
df
          A         B         C
0 -0.467542  0.469146 -0.861848
1 -0.823205 -0.167087 -0.759942
2 -1.508202  1.361894 -0.166701
3  0.394143 -0.287349 -0.978102
4 -0.160431  1.054736 -0.785250

# Select column using a single label, 'A'
df['A']
0   -0.467542
1   -0.823205
2   -1.508202
3    0.394143
4   -0.160431

# Select multiple columns using an array of labels, ['A', 'C']
df[['A', 'C']]
          A         C
0 -0.467542 -0.861848
1 -0.823205 -0.759942
2 -1.508202 -0.166701
3  0.394143 -0.978102
4 -0.160431 -0.785250

其他详细信息请访问： http ： //pandas.pydata.org/pandas-docs/version/0.18.0/indexing.html#selection-by-label