31 thoughts on “R packages”

1. Unfortunately, the current installation code does fail on Macs: here’s its output on R 3.0.0.

   Reply ↓

   • Hmm… seems to be some issue with the rms-package. Have you tried installing without the dependencies?

     Reply ↓

     • It works if you install “rms” separately:

       install.packages("rms")
reps = c("http://ftp.sunet.se/pub/lang/CRAN",
         "http://cran.gforge.se")
install.packages("Gmisc", repos = reps, dependencies = FALSE, type = "source")

       Reply ↓

2. Pingback: Some Results | Shige's Results

3. Hi,

   I am currently playing around with your transition plots and I like it very much.

   Features missing I think are options for setting headers and subtitles as well as enabling the user to give all boxes individual texts. I tried the latter in order to also give an information about reaction times in a pre-post-measurement design, but now I get the error:

   Error: could not find function “getGridVal”

   while I have a help file available on that function, it is neither in my workspace after loading Gmisc, nor can I find in your Gmisc github repository.

   What happended here?

   Reply ↓

   • The latest version allows for creating your own layout around the plot, just push a grid viewport(), plot using the newpage=FALSE into that viewport. Also, see if you still get the error after setting the newpage=TRUE – there may be some old plot info that causes issues. I’m planning on adding the header option, unfortunately I haven’t had time to implement it yet but you can have a look at the forestplot2() that has the header option working in a similar fashion.

     You can put any text you want in your boxes, I’m currently writing an article on Charnley classification A, B and C before and one year after surgery. In the transitionPlot() call I simply have:

     box_txt=cbind(c("Charnley\nclass\nA", "B", "C\nPre-op."), c("Charnley\nclass\nA", "B", "C\nPost-op."))

     Reply ↓

4. Hi Max,

   I really liked ‘transitionPlot’ function in Gmisc package.
   However I would like to know if we can have a transition plot like
   A >> B >> C

   Reply ↓

   • Hi Abhinaba,

     I’ve had this request posted previously but I haven’t had the time to implement it yet. It is not incredibly difficult, but it would require some rethinking. If you want to try to adapt the code you are welcome to write a suggestion for the package. Just fork the Github-repository and write a suggestion. You probably don’t need to get down to the details as the major sections are ready-to-use building blocks. I will try to help you the best I can.

     /Max

     Reply ↓

5. Thanks for all your great efforts on your packages and tutorials. They’re really helpful. However, when I try to run the last R code chunk, I get an error that reads as below and halts execution:

   Error in missing(xlab) : ‘missing’ can only be used for arguments
   Calls: … withCallingHandlers -> withVisible -> eval -> eval -> forestplotRegrObj
   In addition: Warning message:
   In fix_options(params) :
   You must not set both chunk options out.width and fig.retina

   Reply ↓

   • Ok, fixed it. In the version 1.0 I’m trying to make the functions cleaner and I happened to overdo the cleaning. Re-install from Github and see if it works better now.

     Reply ↓

6. Hi Max,
   this package is simply great and I am more than happy for the possibilities it offers to me to turn our data analysis more reproducible and effective by it.
   However, working on it I have one point. I tried to adapt the script for comparing three groups and just replacing the wilcox.test by the kruskal.test works out perfectly.
   But doing the same with the fisher.test – i.d. replacing it by the chisq.test, brings up the error code:
   Error in getDescriptionStatsBy(parameter, group, :
   could not find function “prDescGetAndValidateDefaultRef”
   Looking through the code I found the call for prDescGetAndValidateDefaultRef (line 119), but I couldn’t find the definition for the function.
   Could you help me?
   Best Wishes Aarne

   Reply ↓

   • Sorry, forgot to reply to this comment. It seems to work for me with the 1.1 version – is this still a problem?

     Reply ↓

7. First,

   let me thank you for the forestplot library. It is immensely helpful in creating some logistic regression/Odds ratio plots we needed for publication. That being said, I’m am *very new* to R, and for the life of me cannot figure out how to re-engineer this to change the line style. I’m sure the editor is going to want this in black and white. I have symbols separating my groups (as in your sweden/denmark example from another post), and have changed the colors to Black/Grey to try and differentiate the groups further. But differentiating by line type (say solid vs. dashed) would be great

   But I am simply too new to R to sensibly figure this out. Any possibility of a brief post on how to change the line style itself?

   Again, thank you for a great package.

   Reply ↓

   • Just added this feature to the development version run: devtools::install_github("gforge/forestplot", ref="develop") The line types can be set the same way as any other line-specific parameter through lty.ci. See the vignette for an example.

     I’m not entirely satisfied regarding how to name the argument for the T vertices that mark the ends of the confidence interval, let me know if you have a better naming suggestion. Due to the dashed nature of the line I believe that this necessary in order to avoid any confusion regarding exact line length.

     Reply ↓

     • Thank you ever so much! Works perfectly.

       Reply ↓

8. Hi Max,

   This is great and extremely helpful in my Biostatistics homework, thank you very much. I’m just beginning to learn R.

   I am having an issue with the printCrudeAndAdjustedModel command. It’s giving me HTML output rather than the table. Is there some way I can correct this?

   Reply ↓

9. Hi Max,

   thanks for your valuable contributions! In Gmisc, can you give me hints on how to adapt the style of arrowheads (I basically want no heads, having arrows going in and out just straight)? Thanks!

   Reply ↓

   • Hi Sascha,

     not sure I follow exactly what you want. If you want plain arrows you may be better off with simple grid::polygonGrob implementation than adapting my arrows. The core functionality in the arrows is the Bezier spline that allows you to have a flexible, smooth line. The core idea is to have a center spline and then simply do an offset of x mm in 90 degree to the current segments direction. There are a few things that get complicated (1) if the line is heavily curved then the offset create an inner loop creating unexpected artifacts (2) the arrow is straight and therefore the line mus be shortened in order to compensate for this. If the attaching line isn’t a direct continuation of the arrow it looks rather weird in my opinion. Unfortunately reducing the line forces a recalculation of the spline points which has been a trial and error process where I’ve tested different adaptations to see how close the line would actually follow the original line.

     If you still feel that this is worth-while then you should have a look at the Gmisc:::getBezierAdj4Arrw function and try to generalize it so that you add an arrow at both ends. Under tests/visual_tests/test-bezier-adjustment.R you can find some of the tests that I used when developing the adjustment. The function generates two lines, one left and one right line, that constitute the two borders that together make up the boundaries of the arrow. This is done in the Gmisc:::calculateLinesAndArrow function that you will have to generalize so that it can handle alternative arrow specifications.

     If you get it to work, don’t hesitate to add a pull request into the repository. Good luck!

     Reply ↓

     • Hi Max,

       thanks a lot for your response. However, I got the impression I couldn’t describe my intention properly, which is only about the arrow_head_ instead of the entire arrow (I appreciate the twosided bezier arrows very much). I hope the linked picture clarifies what I desire: http://oc.sascha-holzhauer.de/index.php/s/LmOH2NEj8rlBsj9

       PS: As you see I extended the code for multiple transitions. Unfortuntely, I based my code on Gmisc version 1.0.2 and missed to send a pull request promptly. I hope I find resources to integrate my changes in the current codebase soon.

       Reply ↓

       • A pity that you spent time extending the code for multiple transitions since I’ve already done that. You should now use the TransitionClass, I should post the vignette on my blog but I’ve been too busy.

         Substituting the arrow to a straight line is basically the same set of functions although you don’t need to figure out how to change the Bezier spline. The issue you will have in your example is that not all lines come in at a straight angle and you will have to us the Gmisc:::align2Axis on the arrow side. The devil is in the details…

         Reply ↓

         • Ah, ok. Looks like you refactored a lot. Thanks for linking the vignette. A few remarks: the link to pacakges diagram seems wrong; formating in ‘Basic use’ seems corrupted (regarding subheaders); a continouse scale for male/female is rather funny…

           Another question: have you thought about arrow widths proportional to box height, i.e. if a flow is 20% of a box’ population the arrow width should be a fifth of box height?

10. Thanks for the bug report. Male/Female is the proportion that transitions… perhaps not the most intuitive example.

    Check out the lwd_prop_type argument. It gives you the options of choosing algorithm for arrow width.

    Reply ↓

11. why the following code did not work?
    devtools::install_github(“gforge/forestplot”, ref=”develop”)

    Reply ↓

12. Hi

    I can’t save forestplot. Can you guide to save forestplot?

    Thank you

    Reply ↓

    • There are several ways of doing this, the easiest is probably to use RStudio with the export function

      Reply ↓

13. Thanks for your fantastic forestplot package. What is the syntax to change the shape of a summary row, from diamond (default) to box and whisker (like the default for non-summary rows)?

    Reply ↓

    • Sorry but I don’t have the time to support general usage, try the docs (the vignettes are the best) or StackOverflow.

      Reply ↓

14. Hi Max,
    Glad to have your package in my meta-analysis. However, when I tried to make my plots more visual and colorful by applying the function fpColors( ), I received *Error in fpColors(box = “#660000”) : could not find function “fpColors” *. I supposed that function was included in the package forestploter, wasn’t it ? Is there some way to solve it?

    Reply ↓

    • The new version of the package uses the fp_set_style() function for most of this type of functionality.

      Reply ↓

15. Hi Max

    Thanks for a great package – forestplot. Any way to get a label that explains the direction of the effect? i.e. favors control / favors intervention
    Like this: https://www.metafor-project.org/doku.php/plots:forest_plot_bmj

    Best
    Theis Itenov

    Reply ↓

    • Sorry for not responding – I recommend posting these questions on StackOverflow as I don’t have the time to help everyone out. I don’t think effect direction is implemented yet. It would be nice to have but I am not sure when I’ll have the time adding it to the package.

      Reply ↓

This site uses Akismet to reduce spam. Learn how your comment data is processed.