Generate ExampleSets between 2 ExampleSets in a row and create a Time Series Table

VictorS

Hi all, 

I have a data set named "Data Set" (find attached). I want to transform this Data set in a "time series"-like data and join with another table that is a real times series. 

In the Sheet "Original Data" there are 5 Atts:
- Start time
- End time
- Speed
- Category
- Duration, in minutes.

I combined the "Start time" and "End time" in one Att. "Time" (see Sheet "Transformation_1").
The time indices between 2 ExampleSets are not strictly monotonic increasing. This means that between 3 ExampleSets in a row will be different time stamps. 

Can somebody give a hint how could i generate ExampleSets = with the number of minutes between 2 ExampleSets in a row?
e.g:                    Time                        Speed
ExampleSet1:  01.01.20 04:40      111
ExampleSet2:  01.01.20 10:20           0

Difference between this 2 = 340 Minutes
Between the 2 ExampleSets I need to create 340 new ExampleSets that will have the "Speed" = 111.

It would be a solution similar to what the "Equilize Time Stamp" operator does, but with no monotonic time stamp between examples.  

Thanx in advance for any idea.

yyhuang

Hi @VictorS,

Thanks for sharing the sample data. Do you have experience with "Create ExampleSet" operator for generating new time series data? It will use the start date and intervals to create new examples. You can put it inside a loop for product sessions and stop sessions individually.

Attached is the process for that demo. 

Let me know how it works (or not).

Cheers,
YY

VictorS

Hi YY, 

thank you very much! This is exactly what I was looking for.
Works perfect. 

Regards,

Victor