Flexible Visualization

svjosvjo Member Posts: 4 Newbie
edited August 2019 in Help
Hi there,

I am new to RapidMiner (using Studio version 9.3) and I am struggeling with the visualization capabilities. What I want to achieve as a result is a scatter plot of predicted values vs. real values, which is easily done. Further, I would like to have the same value range for both axes so that the identity line (real values vs. real values) should be 45 degrees. With some manual adjustments I can achieve an acceptable result for that, too. My problem is that I would like to get that kind of resulting chart for each example set within an IOObjectCollection and I don't want to manually adjust the axes and the titel of the chart for each example set. The question is if there is any possibility to work with macros within visualizations, for example to pass a titel based on the selected example set and to pass minimum and maximum values of an attribute based on the selected example set to the axes so that the charts adjust themselves with respect to the input data and do all have the same appearance? I know I could solve this problem using R so I guess it should not be a problem for RapidMiner as well, but I have not found any tutorial or discussion so far which is covering this topic.

In my opinion, with my experience with RM so far, I would find it helpful to have operators creating charts within processes where you can adjust the parameters as you like, use it within loops and export them automatically using export operators.

Thx for your help.

Best regards,

Best Answers


  • Options
    SGolbertSGolbert RapidMiner Certified Analyst, Member Posts: 344 Unicorn
    Hi Sven,

    you can save the chart configuration and import it later on for another visualization:

    The process is not automatic though. RapidMiner visualizations are thought to be used in data exploration mostly, so if you need to automatically generate a large amount of plots, using R/Python is the way to go.

  • Options
    svjosvjo Member Posts: 4 Newbie
    Hi Sebastian,

    thank you for your answer. Saving the chart configuration is not solving my problem though as each chart needs an individual configuration. For example different ranges of values at the axes.
    I am used to the fact that decisionmakers like charts for visualizing the impact of results of data analyses. For that reason and for documentation purposes I do indeed need to generate a large amount of plots from time to time. I agree, that for data exploration purposes the visualizations in RapidMiner are sufficient but nevertheless I am surprised that RM is not able to create flexible charts. To me this is a large disadvantage of RM compared to for example R, especially regarding the costs for a RM license.
    I think operators for creating plots would be very helpful and comfortable and would erase this disadvantage.

Sign In or Register to comment.