Fit Lines and Curves to your Data in Spotfire

แชร์
ฝัง
  • เผยแพร่เมื่อ 10 พ.ย. 2024

ความคิดเห็น • 11

  • @abhiramkukunuri5587
    @abhiramkukunuri5587 3 ปีที่แล้ว +3

    Would you plan to have an option to force the intercept through 0 for a straight line fit ? This is typically available in excel

    • @losezubi9468
      @losezubi9468 3 ปีที่แล้ว +1

      I have asked for this since 2012, I think. It is standard in Excel, not sure why can't be in Spotfire. You should actually be able to regress / intercept to any point in the Y axis, like in Excel.

  • @losezubi9468
    @losezubi9468 4 ปีที่แล้ว +2

    Very useful Video Neil! Do you have or are you planning to make a video on how the regression parameters from the fittings can be extracted into the actual dxp file, either as a new table or columns into an existing Spotfire table, rather than exporting these to csv files and then having to import then back again? Thanks!

    • @losezubi9468
      @losezubi9468 3 ปีที่แล้ว

      @Neil Kanungo Thanks. For users like me, not interested in mastering programming skills, but use Spotfire as a quick tool to analyse and visualise data, this would be a great addition.

  • @mimosa9188
    @mimosa9188 ปีที่แล้ว +1

    hi how to add a 95 confidence band over the linear fit?

    • @DrSpotfire
      @DrSpotfire  ปีที่แล้ว

      you can use the out of the box Lines & Curves > Straight line fit option or you can also calculate the confidence interval for each x value. The formula for this calculation depends on the standard error of the estimate, degrees of freedom, and the critical value for a 95% confidence interval (usually around 1.96). Then you can add additional lines from these custom expressions
      [LinearFitColumn] + 1.96 * [StandardErrorColumn] as UpperConfidence,
      [LinearFitColumn] - 1.96 * [StandardErrorColumn] as LowerConfidence

  • @vairam1977
    @vairam1977 2 ปีที่แล้ว

    Could you please explain how can I plot fit line for median over sigma? Presently we have straight line option which is regression way r2. But I am looking for Median/Sigma fit line. On my plot I used NormInv((RankReal(dataset1]],"desc","ties.method=first") - 0.5) / count(dataset1) to get the plot ready excep the fit line. Please share insight on what equation should I be using to get the median/sigma fit line instead of straight line

  • @tousifahamed11
    @tousifahamed11 4 ปีที่แล้ว

    This is really great

  • @rafaelsuevo9632
    @rafaelsuevo9632 3 ปีที่แล้ว +1

    Good video. I´m trying do add a custom curve on my graph but it keeps "loading" (circle spinnig on the screen) and never shows anything. Eventualy it shows a blurred background on the whole graph. I hate when things work on the video but never in real life!

    • @rafaelsuevo9632
      @rafaelsuevo9632 3 ปีที่แล้ว

      @Neil Kanungo any time, I guess. I'll try to ask people what's going on. Thanks for your suggestion.

  • @CHINTALAGOPI
    @CHINTALAGOPI 3 ปีที่แล้ว

    Is it possible to add different horizontal lines per trellis?