With 1.2 under our belts, we go now to the example in section 1.3 which was designed to show us how to partition a larger set of data into subsets for analysis. In this case, we’re going to jump to example 1.3.2 to determine the number of live births.
While the Python loop is easy to write, the R code is even easier:
livebirths <- subset(pregnancies,outcome==1)
First: don’t let the <- throw you off. It's just a more mathematical presentation of "=" (the assignment operator). While later versions of R support using = for assignment operations, it's considered good form to continue to use the left arrow.
The subset function will traverse pregnancies, looking for fields (variables) that meet the boolean expression outcome == 1 and place all those records into livebirths.
You can apply any amount of field logic to the subset function, as asked for by example 1.3.3:
firstbabies <- subset(pregnancies,birthord==1 & outcome==1)
Since R was built for statistical computing, it's no surprise that to solve example 1.3.4 all we have to do is ask R to return the mean of that portion of the data frame:
mean(firstbabies$prglength)
mean(notfirstbabies$prglength)
(Here's a refresher on the basics of R data frame usage in case you skipped over that URL in the first post.)
To get the ~13hrs difference the text states, it's simple math. Just subtract the two values, multiply by 7 (days in a week) and then again by 24 (hours in a day).
In the next post, we'll begin to tap into the more visual side of R, but for now, play around with the following source code as you finish working through chapter one of Think Stats (you can also download the book for free from Green Tea Press).
# ThinkStats in R by @hrbrmstr# Example 1.3# File format info: http://www.cdc.gov/nchs/nsfg/nsfg_cycle6.htm# setup a data frame that has the field start/end infopFields <- data.frame(name = c('caseid', 'nbrnaliv', 'babysex', 'birthwgt_lb','birthwgt_oz','prglength', 'outcome', 'birthord', 'agepreg', 'finalwgt'),
begin = c(1, 22, 56, 57, 59, 275, 277, 278, 284, 423),
end = c(12, 22, 56, 58, 60, 276, 277, 279, 287, 440)
)# calculate widtds so we can pass them to read.fwf()pFields$width <- pFields$end - pFields$begin + 1
# we aren't reading every field (for the book exercises)pFields$skip <- (-c(pFields$begin[-1]-pFields$end[-nrow(pFields)]-1,0))
widths <- c(t(pFields[,4:5]))
widths <- widths[widths!=0]
# read in the filepregnancies <- read.fwf("2002FemPreg.dat", widths)
# assign column namesnames(pregnancies) <- pFields$name
# divide mother's age by 100pregnancies$agepreg <- pregnancies$agepreg / 100
# convert weight at birth from lbs/oz to total ouncespregnancies$totalwgt_oz = pregnancies$birthwgt_lb * 16 + pregnancies$birthwgt_oz
rFields <- data.frame(name = c('caseid'),
begin = c(1),
end = c(12)
)rFields$width <- rFields$end - rFields$begin + 1
rFields$skip <- (-c(rFields$begin[-1]-rFields$end[-nrow(rFields)]-1,0))
widths <- c(t(rFields[,4:5]))
widths <- widths[widths!=0]
respondents <- read.fwf("2002FemResp.dat", widths)
names(respondents) <- rFields$name
# exercise 1# not exactly the same, but even more info is provided in the summary from str()str(respondents)
str(pregnancies)
# for exercise 2# use subset() on the data frames# again, lazy use of str() for outputstr(livebirths)
livebirths <- subset(pregnancies,outcome==1)
# exercise 3firstbabies <- subset(pregnancies,birthord==1 & outcome==1)
notfirstbabies <- subset(pregnancies,birthord > 1 & outcome==1)
str(firstbabies)
str(notfirstbabies)
# exercise 4mean(firstbabies$prglength)
mean(notfirstbabies$prglength)
hours = (mean(firstbabies$prglength) - mean(notfirstbabies$prglength)) * 7 * 24
hours
