My data files consist of roughly 1 million rows of time-series data. It has been read into Python using df = pd.read_csv(...)
.
I am looking for a way to get the duration of the file (in seconds), the output I am looking for is just one number to give the duration
Below shows the first and last 5 entries to show the structure of the data:
df.head(5)
X Y Z
TimeStamp
2017-05-12 11:03:39.560 0.185310 -0.168226 0.385064
2017-05-12 11:03:39.570 0.184273 -0.290579 0.497026
2017-05-12 11:03:39.580 0.188649 -0.456002 0.601236
2017-05-12 11:03:39.590 0.195188 -0.629775 0.679267
2017-05-12 11:03:39.600 0.196400 -0.789999 0.729308
df.tail(5)
X Y Z
TimeStamp
2017-05-12 13:18:59.950 -0.045288 -0.018508 1.010065
2017-05-12 13:18:59.960 -0.045412 -0.018438 1.009695
2017-05-12 13:18:59.970 -0.045671 -0.018282 1.009768
2017-05-12 13:18:59.980 -0.045889 -0.018029 1.010952
2017-05-12 13:18:59.990 -0.045657 -0.017709 1.013374
IIUC, let's try, given TimeStamp is a DatetimeIndex: First let's get you index into datetime:
df.index = pd.to_datetime(df.index)
df.reset_index()['TimeStamp'].diff().sum().total_seconds()
OR
(df.index[-1] - df.index[0]).total_seconds()