Here are a few notes to get you started in R.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Help %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Type help(XXX) to get help on command XXX
Type help.search("XXX") to get help on topic XXX (not the quotes)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% I/O %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Read data using scan
convert to a matrix with "matrix"

 m <- matrix( scan ("mfile"), ncol = 5, byrow=TRUE)  - convert data to a matrix

To save and restore in R format
  save(x)
  load(x)

To write in readable format
  write(as.matrix(zz), file = "junk")

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% REGRESSION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

I use glm() for regression ("family" parameter specifies the type of
regression, logistic ("binomial"), linear ("gaussian"), etc). To do
standard linear regresion, you can use "ls".

 For step-wise I use "step" where you can specify "forward", "backward" or
"both", the scope, i.e. upper and lower "model boundaries", and penalty
"k" (for BIC it is log(N)).

To find out what one knows about the result "myresult" which is a regression object:
  attributes(myresult)
  myresult$dof.residual

WARNING stepwise regression seems to give errors when a zero row is tried 

names(<object>)   to see what fields it has
<object>$<fieldname> to extract the specified field



%% basic regression - example (executable R code)

nrows <- 100
ncols <- 1000
x <- matrix (rnorm(nrows*ncols), ncol = ncols)
y <- apply(x[,1:3],1,mean) + 0.05 * rnorm(nrows)
traindata <- data.frame(x,y)
glmtest <- glm(y ~ 1,   gaussian(link = identity ), data= traindata) 

summary(step(glmtest, direction="forward", scope = list(lower = ~ 1, 
upper = ~ 1+X1+X2+X3+X4+X5+X6+X7+X8+X9+X10+X11+X12+X13+X14+X15), 
k = log(nrows)))

summary(step(glmtest, direction="forward", scope = list(lower = ~ 1, 
upper = ~ 1+
X1+X2+X3+X4+X5+X6+X7+X8+X9+X10+X11+X12+X13+X14+X15+X16+X17+X18+X19
+X20+X21+X22+X23+X24+X25+X26+X27+X28+X29+X30+X31+X32+X33+X34+X35+X36
+X37+X38+X39+X40+X41+X42+X43+X44+X45+X46+X47+X48+X49+X50+X51+X52+X53
+X54+X55+X56+X57+X58+X59+X60+X61+X62+X63+X64+X65+X66+X67+X68+X69+X70
+X71+X72+X73+X74+X75+X76+X77+X78+X79+X80+X81+X82+X83+X84+X85+X86+X87
+X88+X89+X90+X91+X92+X93+X94+X95+X96+X97+X98+X99+X100), 
k = log(ncols)))

above must be on a single line; no line breaks
log(nrows) gives 7 features in the model
log(ncols) gives 3 features (correctly)