Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Removing outliers from stata boxplot view

    Dear all
    I am using Stata for my master thesis, and have some problems figuring out how to remove the outliers from my boxplot.
    Reason I want to remove the outlier is due to the fact that I use boxplot to display my data graphically, and just want to focus on the quartiles in the main report, as the boxplot with the outlier will be presented in appendix.
    I am able to remove outliers within the graph editor, but this is done one by one, which takes too much time.

    ​Thanks in advance!

    Kind regards,

    Roman

  • #2
    If you collapse to median and quartiles, and then draw box plots, you will get a box plot showing those alone.

    This seems rather extreme, although it has been asked before. I would prefer to find a transformation in most cases.

    See these threads from April

    https://www.statalist.org/forums/for...-from-box-plot

    https://www.statalist.org/forums/for...median-and-iqr

    and know that a Tip on the topic is in press at Stata Journal 19(4) and likely to be published in December.
    Last edited by Nick Cox; 01 Oct 2019, 03:29.

    Comment


    • #3
      Thank you for the answer!
      I found another solution by acutally reading Stata manuel, which stated that you can use command nooutsides, to remove all outliners

      Comment


      • #4
        Indeed, but that leaves whiskers in place -- which didn't seem to be what you were asking.

        I use boxplot to display my data graphically, and just want to focus on the quartiles in the main report

        Comment


        • #5
          Oh yea, sorry I forgot to say, that I actually wanted those in the boxplot aswell, so that command was fine for me

          Another offtopic question, if you read this, that I can't figure out:
          How do I generate a new table, where I want to transform nubmers into text, such as value 2 in one table = Monday in my new generated table. I have values 2,3,4,5,6 - representing all days between monday-friday.

          Comment

          Working...
          X