My dataset of accounting data on 110,000 companies has its primary existence in the form of Excel spreadsheets. Each year's data is stored in a directory containing 13 spreadsheet files (with columns
First, some background: My project involves taking a large number of randomly-sized random samples using the pair of functions appended to this message. As can be seen, RS2.sample() produces a sample
My dataset of accounting data on 110,000 companies has its primary existence in the form of Excel spreadsheets. Each year's data is stored in a directory containing 13 spreadsheet files (with columns
I need to produce a series of samples and save each sample to a file, then recall the samples and do things with them. I've managed to write a script to produce the series of samples and save them, b
I'm trying to do something like foo.list <- list(foo.1, foo.2, ...., foo.n) where foo.n are n vectors There must be a more elegant way to create these n arguments than laboriously typing in "foo.1"..
Last week I asked how one could save oneself the effort of typing out a long series of arguments to list(), when the arguments were in a standardised format, e.g.: foo.list <- list(foo.1, foo.2, ....
There's a limit to the amount of text one can cut-and-paste to the command window (see the thread "Getting a large amount of code from Word to the Command", started by Ken Kelley on 24 May 2001-- htt
The following function takes n = how.many *randomly-sized* random samples from p, in which the probability of selection for p[i] depends on the value of k[i] relative to max(k) -- thus, if max(k) = k
Thanks to Bill Dunlap, Nick Ellis, Steve McKinney, Brian Ripley, Dimitris Rizopoulos and Greg Warnes for their replies. I've learnt a lot from their various approaches (reproduced below); the essenti
My S+ 2000 procedure of the form thing <- foo(x, y) takes about 12 minutes on a Pentium Pro (processor speed not to hand) with 256MB RAM. Since foo() involves calculating the first through fourth L-m
Many thanks to those who replied. Prof Ripley and Michael Prager both suggested cutting and pasting from the script window to the commands window, which seems to do the business -- about 12 minutes f