cisim() needs update like nsize

[alanarnholt]

Alan...update cisim() without using \b in code to work in LaTeX

[ibecav]

OurConfidence.R.txt
Hi Alan,

Thanks for your hard work on this. I also teach statistics occasionally and stumbled across your cisim function. I wanted to make some functional changes like ensuring that if a student ran it they would get back plots that had consistent Y scaling where appropriate so they could see the intervals "grow" and "shrink" as they changed a parameter.

While I was at it I decided I might as well make use of ggplot2 to get a more polished plot.

I'm enclosing the rewritten code. For now I have only worked on the means portion. But if it's useful I could certainly rewrite the rest. On the other hand if you prefer your original code I'll go my own way...

Thanks again.

Chuck

[alanarnholt]

Dear Chuck,

Thanks for the comments. I am certainly open to any improvements and code contributions. I will take a closer look at your proposal later this week. There are a few problems in the text that accompany PASWR2 that ask the students to modify the cisim() function so I am thinking adding an option to use either base graphs or ggplot2 graphs (both with properly scaled Y axis) may be a good approach.

[ibecav]

Alan,

No pressure no worry. Just trying to give back and credit your work.

Chuck

[alanarnholt]

Dear Chuck,

I extended the code you sent me for CIs for a proportion and variance (copied below- still playing with edge cases). I understand the idea of fixing the y axis for the mean so users see the length of the CIs shrink as n increases. The function probably needs an option to use or not use this setting. It becomes challenging to find the "right" limits for the other confidence intervals. I am attaching the code as it stands and a link to a shinyapp where the function is being used. Any comments are most appreciated.

Shinyapp: https://alanarnholt.shinyapps.io/CISIMG/

cisimgg.txt

[ibecav]

Very nice Alan. Glad it was helpful. The Shiny app is very well done.