Best approach to join incomplete time series data

[mi_schmidt]

Hi Alteryx Community,

I'm often challenged with time series data from sensors where the sensors just send a new value when a hange has happend. That results in arbitrarily time intervalls between two values. My goal is to analyze and visualize (in Tableau) the dependency of multiple parameters from different data streams.

So far I joined two of those data streams with a generated list of time steps, e.g. for each minute. Unfortunately, the complexity raises when additional parameters (in this example the Pos) come into account.

Do have any proposal to improve my current approach?

Thanks in Advance for your help

Michael

[echuong1 · Accepted solution]

Have you looked into the TS Filler tool? This will help fill in any gaps within your data. 

https://community.alteryx.com/t5/Alteryx-Designer-Knowledge-Base/Tool-Mastery-TS-Filler/ta-p/84642

[mi_schmidt]

Hi,

Thanks for your quick response. 

I haven't used the TS Fill+er tool yet and will give it a try!

Thanks

Michael

[RolandSchubert]

Hi @mi_schmidt ,

in my understanding, you have different sensors (A, B) at different positions (110, 115, 121) sending a signal, if change hanppens. If that is true, you could use a Multi-Row Formula tool to fill the gaps (sensor B at position 110 as the specific values at 00:26:00 until 01:00:00, this is the next entry, values unchanged, Multi-Row Formula could fill in the values registered at 00:26:00 up to 00:59:00, repeat the process starting at 01:00:00 to the next entry). 

But - to be honest - I was only guessing, so a little bit more backgroupn would be helpful. What do you think?

Best,

Roland 

[mi_schmidt]

Hi @RolandSchubert,

thanks for your help.

Your assumptions are completly right and the solution should work as well. I tried the TS Filler and could achieve the desired results.

Best

Michael