Pandas Data Frame根据列值查找索引

I have a data frame lets say "df". Now one of the columns of the data frame is named "itemID". I would like to get some how very fast the row index according to a value on the column "itemID".

我有一个数据框让我们说“df”。现在，数据框的一列被命名为“itemID”。我想根据列“itemID”上的值得到一些非常快的行索引。

When I do:

当我做：

df[df['itemID']==X]

The performance is quite slow.

表现很慢。

Is there a way to create something like a hash-index in order to do the above?

有没有办法创建类似哈希索引的东西才能完成上述操作？

1 个解决方案

#1

I believe you can use dask.

我相信你可以使用dask。

Docs say:

文件说：

The following class of computations works well:

以下类计算效果很好：

Trivially parallelizable operations (fast):

平凡可并行化的操作（快速）：

Row-wise selections: df[df.x > 0]

行方式选择：df [df.x> 0]

You can also check how Create Dask DataFrames.

您还可以查看Create Dask DataFrames的方式。

Example

例

import pandas as pd
import dask.dataframe as dd

df = pd.DataFrame({'A': ['A0', 'A1', 'A2', 'A3'],
                   'itemID': [1,2,4,4]})

print (df)
    A  itemID
0  A0       1
1  A1       2
2  A2       4
3  A3       4

#Construct a dask objects from a pandas objects
df_dask = dd.from_pandas(df, npartitions=3)

#Row-wise selections
print (df_dask[df_dask.itemID == 4].compute())
    A  itemID
2  A2       4
3  A3       4

#1