search

briandk / granovaGG

Bob Pruzek and Jim Helmreich's implementation of Elemental Graphics for Analysis of Variance
Other
15 stars 4 forks source link

Needed additions/changes to CRAN version of granovaGG 1.0 #118

Closed rmpruzek closed 13 years ago

rmpruzek commented 13 years ago

Re: 1w

  1. re function .1w: re: table that displays numeric results (line 1: group group.mean trimmed.mean) a. Reorder the rows to correspond w/ the ranking of means (and effects) from small to large. b. Revise the title to "Summary statistics by-­group for input data (with means ordered from smaller to larger) c. Add numerics as in granova.1w to include: Summary Statistics as per granova.1w, including p-­‐value for F, etc... see $grandsum: $grandsum Grandmean df.bet df.with MS.bet MS.with F.stat F.prob SS.bet/SS.tot 
 Alternatively, could use Summary of aov run as in  aov.obj = aov( response ~ factor(groups), data ...) (2 digit display)
 (misses the last term above, but the rest would be given)

... Further changes, that my various trials suggest, for 1w:

  1. get rid of F = value below plot that is part of xlab; redundant w/ F above ‘box’
  2. make grand mean point (dot) larger, ... approximately double its current diameter
  3. make it possible (new arg needed) to suppress box for F (for descriptive work, e.g. singer data) [not critical]
  4. re function .ds:
    1. For ds the ‘sw’ (used in example) does not seem to be documented (southwest?? Also, setting sw to different values seems to have no effect. I tried (.1 … 6)... all the same plots
    2. Need to add Pruzek, Helmreich (2009) J.S.E. reference (see granova.ds)
    3. None of the stats are printed in my runs, see granova.ds . Please add, as in that function.
    4. Also, I will add one or two+ examples/datasets (Snedecor tobacco leaf, Box, Hunter, Hunter Shoes data) to the package ...4 is because it would be most helpful if there were 4 different examples, one (at least) each for the 4 paradigms in P-H article.
  5. re: function .contr
  6. In several trials, after I enter the run command (e.g. granovagg.contr(rat[,1],contr.helmert(6)) ) even when I don’t touch the keyboard the first 4 panels are printed, then the last two. If I exit R, reload and go … then I get the first four and can press return to get others. BUT then when I try to rerun, the initial problem recurs.
    1. After getting no INITIAL graphics (as it goes to the last page (with 2 panels here)) I see what I pointed out in our late testing sessions, I see “Since you elected to print four plots per page granovagg.contr won't return any plot objects.” Certainly this sentence needs to be completely revised since since any run will yield at least two plots. Curiously, and Jim may have to weigh in this, granova.contr has exactly the same problem, so the issue goes back one generation … and it could be there is some more basic function in R itself. ?? bp
briandk commented 13 years ago

I'm closing this issue because it's been broken out into several smaller issues:

rmpruzek commented 13 years ago

Brian, Your listing here seems to cover the bases, and I'll be interested to see each item implemented. Can you give me any sense of when you see this as likely to be finished given your other interests and obligations? I'd appreciate an email re: this note on this point asap. Thank you, bob p.s. Given that you (and Wil) are listed as coauthors for the presentation I'll be making on or near to Oct. 13, I wanted you to know that I'm putting that together, but will not be likely to have anything to until about the middle of next week. Then I would need rather quick turnaround. I realize that until you see my stuff you will not be able to tell me whether you can provide any response, but in case you could block off an hour or two between Wed, say, and the 10th, this could be helpful.

From: Brian A. Danielak reply@reply.github.com To: rmpruzek rmpruzek@yahoo.com Sent: Friday, September 30, 2011 7:52 PM Subject: Re: [granovaGG] Needed additions/changes to CRAN version of granovaGG 1.0 (#118)

I'm closing this issue because it's been broken out into several smaller issues:

Reply to this email directly or view it on GitHub: https://github.com/briandk/granovaGG/issues/118#issuecomment-2254923

briandk commented 13 years ago

I've been working pretty hard to cover these. Many of the minor fixes have already been implemented:

Two fixes you've asked for are a bit more involved. I've submitted pull requests so people can review the changes. Please provide your input on:

  1. The printed output for granovagg.ds
  2. Suppressing visual squares in granovagg.1w

You can provide input by clicking the links above and offering your comments in the discussion.

rmpruzek commented 13 years ago

Thanks Brian, I've submitted comments on the two URLs you provide. Best, bob

From: Brian A. Danielak reply@reply.github.com To: rmpruzek rmpruzek@yahoo.com Sent: Sunday, October 2, 2011 12:26 AM Subject: Re: [granovaGG] Needed additions/changes to CRAN version of granovaGG 1.0 (#118)

I've been working pretty hard to cover these. Many of the minor fixes have already been implemented:

Two fixes you've asked for are a bit more involved. I've submitted pull requests so people can review the changes. Please provide your input on:

  1. The printed output for granovagg.ds
  2. Suppressing visual squares in granovagg.1w

You can provide input by clicking the links above and offering your comments in the discussion.

Reply to this email directly or view it on GitHub: https://github.com/briandk/granovaGG/issues/118#issuecomment-2260468