What inspired this wish list request?
Our company uses JMP’s PCA charts to monitor high-dimension(level?) data trends in equipment or process conditions.
For example, in the case of Major centrifugal compressors, we collect between 50 to 100 real-time operational variables such as RPM, power consumption, inlet/outlet pressure, flow rates, vibration levels, and data from associated equipment.
We use PCA charts to classify operational states into “Good Condition” and “Bad Condition,” enabling us to track whether the current state is trending toward a problem. In doing so, it is crucial for us to examine which variables are influencing the movement of the principal component scores, by reviewing the eigenvectors (loading vectors).
However, we have recently discovered that JMP Live restricts the labeling of variables on the PCA loading plot to 30 variables or fewer, which limits our ability to monitor the full range of contributing factors. This constraint is inconvenient and hinders the depth of our analysis.
What is the improvement you would like to see?
We would like to see the ability to increase the number of labeled variables displayed on the PCA plot beyond the current 30-variable limit. This would allow us to quickly and directly observe the influence of a wider set of variables on the principal components.
Why is this idea important?
This feature is critical to how we utilize JMP Live in our operations. Solving this limitation would greatly enhance our ability to apply PCA monitoring to various tasks more effectively. It would improve usability, increase analytical transparency, and allow for broader adoption across similar use cases within our organization.