Avatar

Jim Thomas

  • Total activity 104
  • Last activity
  • Member since
  • Following 0 users
  • Followed by 0 users
  • Votes 20
  • Subscriptions 38

Activity overview

Latest activity by Jim Thomas
  • Avatar

    Jim Thomas commented,

    Jay, Just a quick followup.  I was playing with this today and it really does act strangely.  I thought the tipping point where the regression looks incorrect was at 8 units, but with different uni...

  • Avatar

    Jim Thomas commented,

    Jay, Any thoughts on what the issue is here? Regards, Jim

  • Avatar

    Jim Thomas commented,

    Jay, Have been working some more with the Power Regression fit.  I've noticed that as the sample size drops off the result is very skewed from what we would expect.  In our example, where we are lo...

  • Avatar

    Jim Thomas commented,

    Ronen, An improvement.  Even better would be the basing on the values of the intermediate nodes at each level as well so I would know which branch to expand to see identify the issues.  I'm experim...

  • Avatar

    Jim Thomas commented,

    Ronen, Very neat.  The ability to do conditional formatting of nodes based on the value to highlight the outliers or issues would be fantastic. Regards, Jim

  • Avatar

    Jim Thomas commented,

    Jay, Looks perfect. I should get a straight line on a log-log scale and that is exactly what I get. My customer will be very happy. Thank you, Jim

  • Avatar

    Jim Thomas commented,

    Jay, Yes.  That's it.  Thanks,  Jim

  • Avatar

    Jim Thomas commented,

    Jay, Any joy with the Power Law regression?  Our customer is coming in next week.  I would love to be able to show this to him. Thanks, Jim

  • Avatar

    Jim Thomas commented,

    Jay, The regression script library listed a Power Law regression.  That is what we are looking for.  Can you enable it as a menu choice in the plugin? Thanks, Jim

  • Avatar

    Jim Thomas commented,

    Is the Log regression in the menu the same as Excel's Power Model regression?   We have a customer asking for a Power Model regression.