Performance alteryx

[Surname]

HI there, 

does anybody know what is actually the max performance by alteryx? Can I run Excel calculation with e. g. 500 K of data set? 

Thank you in advance!

Regards

Dilyana

[s_pichaipillai]

Hi Dilyana

yes you can run it 

are you running into any issue ?

[Surname]

YeHi, thank you for your reply!

Do you know what are the board of performance? How many data ( e.g. Excel data) can I survyve with alteryx? 

Thank you!

[s_pichaipillai]

Performance wise Alteryx does good job. Even i have used Alteryx to process Millions of rows in less time?

[Surname]

How many time did it get? 

Usually in excel you need one day in order to do calculation with such data volumen.

[s_pichaipillai]

well, Alteryx has so many tools inside and you can use them for your calculation etc....

Right now i dont have any idea about your full requirement, so if you could share a sample data and what are you trying to achive, we can help :)

thanks

Saravanan

[dataMack · Accepted solution]

Alteryx is an in-memory tool, whose perfromance is going to vary based on the machine on which you are running it.  I have a somewhat standard corporate laptop thats a few years old (i5 processor, 16GB ram) and I rotinuely throw datasets with tens of millions of records at Alteryx and it runs fast without issue.

Perfromance is also going to vary based on what actions you are trying to perform on that dataset (which tools are part of your workflow).  For instance, a 500k row dataset to which you are doing basic dataprep (summary, filtering, joining) may actually run faster than a 50k row data set that you are running through predictive modeling tools.

[jdunkerley79 · Accepted solution]

I did some processing with the entire Land Price Registry data set in the UK (parsing from CSV and creating Medians and perctiles by post code) which is 20mm row dataset.

I am running a 4th gen i7 with 8gb RAM and SSD (which matters as it will swap out to disk as it needs) and it handled that process in about 4 minutes.

The core engine that @ned and his team wrote, is very performant (doubt many programmers could create faster dedicated code) and multi-million row record sets are no bother to it.