I work in manufacturing and would like to generate pareto's for equipment that show top downtime causes over a period of time using alarm tags. Is this possible to do in Pi?
I would use event frames to prep the data and then PI OLEDB Enterprise (maybe even the new SQL client?) or the Pi Integrator for business analytics and use a BI tool like Power BI or Tableau. PI Vision isn't really set up for this type of investigation.
It is great for time series data, seeing how variables do over events/before events, and some light event comparisons (never been crazy about how that was implemented). BI tools are much better at Pareto analysis.
PI Vision custom symbol can be used to show pareto chart. For example, high chart library contains good example.
If you have coding knowledge, then you can implement it. (Of cause PI Tag or AF Attribute should contain these data first)
Any luck actually implementing one Kenji?
Between the osi timeseries demo and the highcharts docs, it seems like this would be pretty straightforward.
However, I'm currently dealing with some deep JS minification element throwing as undefined which is preventing the entire symbol from rendering.
Highcharts looks to have the best pareto module OOB but I've yet to see any working examples. If anyone knows of any or even has a modified dataUpdate function that produces a highcharts pareto, I'm very interested in checking that out!
I have not tried this pareto chart yet but I implemented the other high chart library. So if you show an error then PI Square community people or I can help you.
It is good to use debug option. In the PI Vision web.config, you can set debug=true and it allows us to do debug. please refer to user guide.
PI Vision 2017 R2 SP1 Extensibility Guide - (English)
<compilation debug="true" targetFramework="4.6"/>
Retrieving data ...