Select sub_Dataframe with postive element and return index in sub

I have a dataframe looking like this.

                         A        
 2013-01-05 00:00:00     0
 2013-01-05 01:00:00     0
 2013-01-05 02:00:00     5
 2013-01-05 03:00:00     20
 2013-01-05 04:00:00     10
 2013-01-05 05:00:00     0
 2013-01-05 06:00:00     0
 2013-01-05 07:00:00     3
 2013-01-05 07:00:00     6

I tried to select sub dataframes with positive values and extract their indexes

List= df[df['A']>0].index.tolist()

Index of first and last positive element of each sub dataframe, put the sub list in list: for this dataframe [[5,10],[3,6]] and return their indexes

Desired output:List[[ 2013-01-05 02:00:00,2013-01-05 04:00:00],[2013-01-05 07:00:00,2013-01-05 08:00:00]]

CodePudding user response：

You could try:

idx_list = (
    df
    .assign(
        group=df["A"].gt(0).diff().fillna(False).cumsum(), idx=df.index
    )[df["A"].gt(0)]
    .groupby("group").agg({"idx": lambda col: [col.iat[0], col.iat[-1]]})
    .idx.to_list()
)

Result with

df = 
                     A
2013-01-05 00:00:00   0
2013-01-05 01:00:00   0
2013-01-05 02:00:00   5
2013-01-05 03:00:00  20
2013-01-05 04:00:00  10
2013-01-05 05:00:00   0
2013-01-05 06:00:00   0
2013-01-05 07:00:00   3
2013-01-05 08:00:00   6

[[Timestamp('2013-01-05 02:00:00'), Timestamp('2013-01-05 04:00:00')],
 [Timestamp('2013-01-05 07:00:00'), Timestamp('2013-01-05 08:00:00')]]

CodePudding user response：

Change index for the column name and then extract the first and the last element from a list:

my_list = df[df['A']>0]['A'].to_list()
my_list= [my_list[0],my_list[-1]]