Sampling and point
estimates
SAMPLING IN R
Richie Cotton
Data Evangelist at DataCamp
Estimating the population of France
A census asks every household how many
people live there.
SAMPLING IN R
There are lots of people in France
Censuses are really expensive!
SAMPLING IN R
Sampling households
It's cheaper to ask a small number of
households and use statistics to make
estimates about the whole population.
Working with a subset of the whole population
is called sampling.
SAMPLING IN R
Population vs. sample
The population is the complete dataset.
It doesn't have to refer to people.
You typically don't know what the whole population is.
The sample is the subset of data you calculate on.
SAMPLING IN R
Coffee rating dataset
total_cup_points variety country_of_origin aroma flavor aftertaste body balance
90.58 NA Ethiopia 8.67 8.83 8.67 8.50 8.42
89.92 Other Ethiopia 8.75 8.67 8.50 8.42 8.42
... ... ... ... ... ... ... ...
73.75 NA Vietnam 6.75 6.67 6.5 6.92 6.83
Each row represents 1 coffee.
1138 rows.
We'll treat this as the population.
SAMPLING IN R
Points vs. flavor: population
pts_vs_flavor_pop <- coffee_ratings %>% total_cup_points flavor
select(total_cup_points, flavor) 1 90.58 8.83
2 89.92 8.67
dim(pts_vs_flavor_pop) 3 89.75 8.50
4 89.00 8.58
... ... ...
1338 2
1335 78.08 7.67
1336 77.17 7.33
1337 75.08 6.83
1338 73.75 6.67
SAMPLING IN R
Points vs. flavor: 10 row sample
pts_vs_flavor_samp <- coffee_ratings %>% total_cup_points flavor
select(total_cup_points, flavor) %>% 1 82.25 7.58
slice_sample(n = 10) 2 83.50 7.67
3 80.50 7.17
dim(pts_vs_flavor_samp) 4 79.33 7.17
5 83.83 7.58
6 84.17 7.75
10 2
7 83.67 8.17
8 81.92 7.50
9 82.67 7.58
10 83.42 7.67
SAMPLING IN R
Base-R sampling
Use slice_sample() for data frames, and sample() for vectors.
cup_points_samp <- sample(coffee_ratings$total_cup_points, size = 10)
88.25 83.83 83.17 82.67 84.67 83.42 73.67 86.00 81.58 80.92
SAMPLING IN R
Population parameters & point estimates
A population parameter is a calculation made on the population dataset.
mean(pts_vs_flavor_pop$total_cup_points)
82.15
A point estimate or sample statistic is a calculation made on the sample dataset.
mean(cup_points_samp)
82.82
SAMPLING IN R
Point estimates with dplyr
pts_vs_flavor_pop %>%
summarize(mean_flavor = mean(flavor))
mean_flavor
1 7.526
pts_vs_flavor_samp %>%
summarize(mean_flavor = mean(flavor))
mean_flavor
1 7.716
SAMPLING IN R
Let's practice!
SAMPLING IN R
Convenience
sampling
SAMPLING IN R
Richie Cotton
Data Evangelist at DataCamp
The Literary Digest election prediction
Prediction: Landon gets 57%; Roosevelt gets 43%
Actual results: Landon got 38%; Roosevelt got 62%
Sample not representative of population, causing sample bias.
Collecting data by the easiest method is called convenience sampling.
SAMPLING IN R
Finding the mean age of French people
Survey 10 people at Disneyland Paris.
Their mean age is 24.6 years.
Will this be a good estimate for all of
France?
1 Image by Sean MacEntee
SAMPLING IN R
How accurate was the survey?
Year Average French Age 24.6 years is a poor estimate.
1975 31.6 People who visit Disneyland aren't
1985 33.6 representative of the whole population.
1995 36.2
2005 38.9
2015 41.2
SAMPLING IN R
Convenience sampling coffee ratings
coffee_ratings %>%
summarize(mean_cup_points = mean(total_cup_points))
mean_cup_points
1 82.09
coffee_ratings_first10 <- coffee_ratings %>%
slice_head(n = 10)
coffee_ratings_first10 %>%
summarize(mean_cup_points = mean(total_cup_points))
mean_cup_points
1 89.1
SAMPLING IN R
Visualizing selection bias
coffee_ratings %>%
ggplot(aes(x = total_cup_points)) +
geom_histogram(binwidth = 2)
coffee_ratings_first10 %>%
ggplot(aes(x = total_cup_points)) +
geom_histogram(binwidth = 2) +
xlim(59, 91)
SAMPLING IN R
Visualizing selection bias 2
coffee_ratings %>%
ggplot(aes(x = total_cup_points)) +
geom_histogram(binwidth = 2)
coffee_ratings %>%
slice_sample(n = 10) %>%
ggplot(aes(x = total_cup_points)) +
geom_histogram(binwidth = 2) +
xlim(59, 91)
SAMPLING IN R
Let's practice!
SAMPLING IN R
Pseudo-random
number generation
SAMPLING IN R
Richie Cotton
Data Evangelist at DataCamp
What does random mean?
{adjective} made, done, happening, or chosen without method or conscious decision.
Oxford Languages
SAMPLING IN R
True random numbers
Generated from physical processes, like flipping coins.
Hotbits uses radioactive decay.
[Link] uses atmospheric noise.
Available in R via the random package.
True randomness is expensive.
1 [Link] 2 [Link]
SAMPLING IN R
Pseudo-random number generation
Next "random" number calculated from seed <- 1
previous "random" number. calc_next_random(seed)
The first "random" number calculated from
3
a seed.
If you start from the same seed value, all calc_next_random(3)
future random numbers will be the same.
2
calc_next_random(2)
SAMPLING IN R
Random number generating functions
function distribution function distribution function distribution
rbeta Beta rgeom Geometric rsignrank Wilcoxon signed rank
rbinom Binomial rhyper Hypergeometric rt t
rcauchy Cauchy rlnorm Lognormal runif Uniform
rchisq Chi-squared rlogis Logistic rweibull Weibull
rexp Exponential rnbinom Negative binomial rwilcox Wilcoxon rank sum
rf F rnorm Normal
rgamma Gamma rpois Poisson
SAMPLING IN R
Visualizing random numbers
rbeta(5000, shape1 = 2, shape2 = 2)
[1] 0.2788 0.7495 0.6485 0.6665 0.6546 0.1575
...
[4996] 0.84719 0.35177 0.92796 0.67603 0.53960
randoms <- [Link](
beta = rbeta(5000, shape1 = 2, shape2 = 2)
)
ggplot(randoms, aes(beta)) +
geom_histogram(binwidth = 0.1)
SAMPLING IN R
Random numbers seeds
[Link](20000229) [Link](20000229)
rnorm(5) rnorm(5)
-1.6538 -0.4028 -0.1654 -0.0734 0.5171 -1.6538 -0.4028 -0.1654 -0.0734 0.5171
rnorm(5) rnorm(5)
1.908 0.379 -1.499 1.625 0.693 1.908 0.379 -1.499 1.625 0.693
SAMPLING IN R
Using a different seed
[Link](20000229) [Link](20041004)
rnorm(5) rnorm(5)
-1.6538 -0.4028 -0.1654 -0.0734 0.5171 -0.6547 -0.7854 -0.0152 0.1514 0.5285
rnorm(5) rnorm(5)
1.908 0.379 -1.499 1.625 0.693 0.748 0.974 0.174 -0.781 -0.930
SAMPLING IN R
Let's practice!
SAMPLING IN R
