有没有办法在 pandas 中传播或 pivot df 但保持/填充空值？

Question

This might be a complicated question, and I am very new to pandas. I basically want to use the script from a machine that can be used as a key for data analysis.

I start with the following data frame:

```
Function       Value
6      WAIT        60.0
7       ACT          27
8      WAIT        30.0
10     WAIT        30.0
12     WAIT        30.0
14     WAIT        30.0
16     WAIT        60.0
18     WAIT        60.0
20     WAIT        60.0
22     WAIT        60.0
24     WAIT        60.0
26     WAIT        60.0
28     WAIT        60.0
30     WAIT        60.0
33      ACT           0
34     WAIT        30.0
36     WAIT        30.0
38     WAIT        90.0
40     WAIT       120.0
42     WAIT       120.0
44     WAIT       120.0
46     WAIT       120.0
```

I want to end up with a data frame that looks like this:

Time_Min  Condition
0        1.0          0
1        1.5         27
2        2.0         27
3        2.5         27
4        3.5         27
5        4.5         27
6        5.5         27
7        6.5         27
8        7.5         27
9        8.5         27
10       9.5         27
11      10.5         27
12      11.0         27
13      11.5          0
14      12.0          0
15      13.5          0
16      15.5          0
17      17.5          0
18      19.5          0
19      21.5          0

Where "WAIT" has been converted to accumulated time in minutes (currently it's seconds between samples) and ACT is 0 when it is not set, and then when it is set to switch to 27, it stays 27 at each time point until it changes.

What's the best way to do this? I want to be able to import scripts that vary in ACT value and have different sampling times, but always produces the same structure of output:)

Answer 1

Time_Min is just the cumsum() of Value divided by 60:

df['Time_Min'] = df['Value'].cumsum().div(60)

Condition is more involved but can be done with mask() and groupby() . Create a boolean mask where Function == ACT , then set non- ACT rows to 0 and cumsum() the ACT groups:

act = df['Function'].eq('ACT')
df['Condition'] = df['Value'].mask(~act, 0).groupby(act.cumsum()).cumsum()

Output:

   Function  Value  Time_Min  Condition
6      WAIT   60.0      1.00        0.0
7       ACT   27.0      1.45       27.0
8      WAIT   30.0      1.95       27.0
10     WAIT   30.0      2.45       27.0
12     WAIT   30.0      2.95       27.0
14     WAIT   30.0      3.45       27.0
16     WAIT   60.0      4.45       27.0
18     WAIT   60.0      5.45       27.0
20     WAIT   60.0      6.45       27.0
22     WAIT   60.0      7.45       27.0
24     WAIT   60.0      8.45       27.0
26     WAIT   60.0      9.45       27.0
28     WAIT   60.0     10.45       27.0
30     WAIT   60.0     11.45       27.0
33      ACT    0.0     11.45        0.0
34     WAIT   30.0     11.95        0.0
36     WAIT   30.0     12.45        0.0
38     WAIT   90.0     13.95        0.0
40     WAIT  120.0     15.95        0.0
42     WAIT  120.0     17.95        0.0
44     WAIT  120.0     19.95        0.0
46     WAIT  120.0     21.95        0.0