3 ms·

The plots and chi-square tests are a single function each, it's the data preparation that takes up so much space. Alas, most statistical analyses involve a lot

by liebke 17y ago

The plots and chi-square tests are a single function each, it's the data preparation that takes up so much space. Alas, most statistical analyses involve a lot of preparatory steps, which are rarely shown in the final write-up.

earl 17y ago
Since I recently did something similar in R: d <- read.csv(file='~/stuff/earlh/Iran_2009.csv', header=T, sep=',') lastDigit <- function(v){ v - 10*floor(v/10) } digits <- lastDigit( c(d$Ahmadinejad, d$Karroubi, d$Mousavi, d$Rezaee)) hist(digits, breaks=10) #chi2 gof tab <- table(digits) n <- length(digits) model <- chisq.test(x=tab, p=rep(0.1, 10)) model # hand generated -- check our work above ts <- 0 for(i in 1:length(tab)){ ts <- ts + ( tab[[i]] - 0.1*n)^2 / (0.1*n) } qchisq(p=1-0.076, df=9) and to be more specific: (def regions (sel votes :cols "Region")) (def ahmadinejad-votes (sel votes :cols "Ahmadinejad")) (def mousavi-votes (sel votes :cols "Mousavi")) (def rezai-votes (sel votes :cols "Rezai")) (def karrubi-votes (sel votes :cols "Karrubi")) # -or- attach(d) (def ahmadinejad (map first-digit ahmadinejad-votes)) (def mousavi (map first-digit mousavi-votes)) (def rezai (map first-digit rezai-votes)) (def karrubi (map first-digit karrubi-votes)) # -or- digits <- lastDigit( c(d$Ahmadinejad, d$Karroubi, d$Mousavi, d$Rezaee)) # not even attached -- could drop d$ etc. While it's obviously a matter of taste, it looks horridly verbose.
- Confusion 17y ago
  Don't mistake explicitness for verbosity. Soon we'll see someone posting a Perl oneliner to do the same, claiming R is also verbose. Conciseness is not a goal, clarity is. if you have to expand the operations in your mind, you may well spend more time thinking about your code then when it's written fully, perhaps factored into a properly named function.
- pygy 17y ago
  It could be done more concisely by using data structures and the (map fn struct) function, but it was expanded for clarty.

by liebke 17y ago

earl 17y ago

Since I recently did something similar in R: d <- read.csv(file='~/stuff/earlh/Iran_2009.csv', header=T, sep=',') lastDigit <- function(v){ v - 10*floor(v/10) } digits <- lastDigit( c(d$Ahmadinejad, d$Karroubi, d$Mousavi, d$Rezaee)) hist(digits, breaks=10) #chi2 gof tab <- table(digits) n <- length(digits) model <- chisq.test(x=tab, p=rep(0.1, 10)) model # hand generated -- check our work above ts <- 0 for(i in 1:length(tab)){ ts <- ts + ( tab[[i]] - 0.1*n)^2 / (0.1*n) } qchisq(p=1-0.076, df=9) and to be more specific: (def regions (sel votes :cols "Region")) (def ahmadinejad-votes (sel votes :cols "Ahmadinejad")) (def mousavi-votes (sel votes :cols "Mousavi")) (def rezai-votes (sel votes :cols "Rezai")) (def karrubi-votes (sel votes :cols "Karrubi")) # -or- attach(d) (def ahmadinejad (map first-digit ahmadinejad-votes)) (def mousavi (map first-digit mousavi-votes)) (def rezai (map first-digit rezai-votes)) (def karrubi (map first-digit karrubi-votes)) # -or- digits <- lastDigit( c(d$Ahmadinejad, d$Karroubi, d$Mousavi, d$Rezaee)) # not even attached -- could drop d$ etc. While it's obviously a matter of taste, it looks horridly verbose.

Confusion 17y ago
Don't mistake explicitness for verbosity. Soon we'll see someone posting a Perl oneliner to do the same, claiming R is also verbose. Conciseness is not a goal, clarity is. if you have to expand the operations in your mind, you may well spend more time thinking about your code then when it's written fully, perhaps factored into a properly named function.
pygy 17y ago
It could be done more concisely by using data structures and the (map fn struct) function, but it was expanded for clarty.