Turtle
  1. Go to...
    1. Turtle's home page
    2. Stats and Data with R
    3. Biometry
    4. Data Analysis
    5. Further information
      1. C.V., etc.
      2. System dynamics
      3. JavaScript
      4. Sinai
      5. Java/JavasScript course
      6. C/C++ course
      7. Costa Rica

  1. PmWiki...
    1. WikiSandbox
    2. Basic Editing
    3. Documentation Index
    4. Cookbook (addons)
    5. installation problems


contact: yc@umn.edu




Part I Data in statistics and R

  • 1. Basic R
  • 2. Data in statistics and in R
  • 3. Presenting data

Part II Probability, densities and distributions

  • 4. Probability and random variables
  • 5. Discrete densities and distributions
  • 6. Continuous distributions and densities
  • 7. The normal and sampling densities

Part III Statistics

  • 8. Exploratory data analysis
  • 9. Point and interval estimation
  • 10. Single sample hypotheses testing
  • 11. Power and sample size for single samples
  • 12. Two samples
  • 13. Power and sample size for two samples
  • 14. Simple linear regression
  • 15. Analysis of variance
  • 16. Simple logistic regression
  • 17. Application: the shape of wars to come



« October 2009 period · December 2009 period »

November 2009
SunMonTueWedThuFriSat
01020304050607
08091011121314
15161718192021
22232425262728
2930     

Calendar:

  • No entries for November 2009.
Printable View

!!General comments, suggestions, corrections, etc. about the book

Entries are sorted from most to least recent.

yc? — 20 May 2008, 10:50

A test message.

ma478zda? — 24 August 2008, 10:49

c345t

Douglas Bates — 11 December 2008, 17:50

Overall I like this chapter very much. There are a couple of peculiarities that I did notice. On page 11 the call

seq(1:10)

is used. That is a peculiar construction, especially to show to a beginner. The : operator produces a sequence so it is a synonym for a typical usage of the function seq. That is

1:10

is equivalent to

seq(1, 10)

or

seq(1, 10, by = 1)

Writing seq(1 : 10) is applying the seq function twice in a rather obscure way. Much better to just use 1:10

Douglas Bates — 11 December 2008, 18:00

In the discussion about source and sink in section 1.1.6 it may be worthwhile noting that the R versions for Windows and Mac OS X contain a text editor with the convenient feature that you can send individual lines or groups of lines to the console window for evaluation. Use File | New script to open a new text file and add the lines shown on page 12. Then move the cursor to the top line and type <ctrl>-R. That sends the current line to the interpreter and moves to the next line. This often comes as a revelation to students when I am teaching about R.

Douglas Bates — 11 December 2008, 21:28

I think that the no.dimnames function shown on pages 32-33 may be more confusing than illuminating for a beginner. It is not easy to grasp what is going on when the index l is incremented inside the indexing expression. If I were going to write the function using a loop like this I would write

dd <- dim(a) for (i in seq_along(dd)) dimnames(a)i? <- rep('', dd[i])

In most cases, if I want to produce a list I use lapply for which the result is always a list but whose first argument doesn't have to be a list. Its a bit obscure but you could generate a list from the dim(a) vector by

dimnames(a) <- lapply(dim(a), rep.int, x = '')

A slightly less obscure version is

dimnames(a) <- lapply(dim(a), function(n) rep('', n))

Y Cohen? — 13 December 2008, 15:45

Douglas: all you points are well taken; I will get to these soon. In the meantime, if you want, I can create a group of instructors, give you a password for it and then create topics that can be updated like: quizzes, exams, lecture notes and misc. Then all the group members can upload files on these topics and share comments. Does this sound reasonable?

(:commentboxchrono:)

Page last modified on December 13, 2008, at 03:45 PM