使用切片和列表索引数据帧

Question

Is it possible to index a data frame using something like

df[[0:12, 14, 19], [0:4 , 6]]

where I want to get rows 0 - 11, 14, 19 and columns 0 - 4, 6. Using iloc I haven't been able to combine slicing along with specific row or column numbers.

Answer 1

将切片和列表一起使用可能是一种更正式的方法，但是将切片转换为列表并将其与单个列/行的列表组合起来可能是最简单的方法。

df.iloc[range(0,12) + [14, 19], range(0,5) + [6]]

Answer 2

It would not allow indexing the dataframe like that, but you can create two lists. One list will have the indeces for the rows that you want and the other for the columns. Then, you can use the loc function to get your result.

The lists are created with the specifics in the example. List comprehension was used for the successive numbers and a list for the individual numbers outside of a range.

list_of_rows=[i for i in range(12)]+[14,19]
list_of_columns=[i for i in range(4)]+[6]
df1=df.loc[list_of_rows,list_of_columns]

Answer 3

I think you can seperate the iloc comand in row 0-11 and column 0-4, for example

data.iloc[0:14, 0:4]

and then do the same for row 0-11 and column 6 and so on, afterwards you can combine the different ilocs