sample()

The sample() function will take a random sample of rows from the data frame, without replacement (of the specified sample size). 

For random sampling with replacement, see the resample() function.

Example 1:

You can take a closer look at how the sample() function works by looking at a couple of examples using a small sample size. In the example code below, you will see how you can take a random sample of rows (without replacement) for a single variable in the row, or for the whole row.

# You can sample rows (without replacement) for a specific variable

# Try sampling 5 Thumb lengths from Fingers a few times

# Notice that you get various combinations of the sample values

print("Random sample of Thumbs, N=5")

sample(Fingers$Thumb, 5)

sample(Fingers$Thumb, 5)

sample(Fingers$Thumb, 5)

# Or you can sample the whole row

# Try sampling 3 whole rows from Fingers a couple of times

print("Sampled Fingers, N=3")

sample(Fingers, 3)

sample(Fingers, 3)

Example of output from running the code above (your values may vary due to sampling variation):

Example 2:

What happens if you take a sample greater than the number of rows in the data? Take a look with this small dataset.

# A small vector with 5 heights (in inches)

Height_in <- c(63, 66, 64, 70, 60)

print("Heights in inches")

Height_in

# Try sampling from the vector a few times

# Notice that you get various combinations of the sample values (without replacement of values)

print("Randomly Sampled Heights, N=5")

sample(Height_in, 5)

sample(Height_in, 5)

sample(Height_in, 5)

# What happens if you take a sample greater than the number of values in the sample?

print("Randomly Sampled Heights, N=10")

sample(Height_in, 10)

sample(Height_in, 10)

sample(Height_in, 10)

Example of output from running the code above (your values may vary due to sampling variation):

Example 3:

Here is an example of how to construct a sampling distribution of means using sample(). It is demonstrated in the context of the Martin v. Westvaco scenario (Martin sued his employer, Westvaco, for age discrimination).

# Save the ages of everyone in Martin's department into an object

everyone <- c(25, 33, 35, 38, 48, 55, 55, 55, 56, 64)

# Generate 1000 samples of 3 people to be randomly "fired" and calculate the mean age of each sample

# Save them into a data frame

SDoM <- do(1000) * mean(sample(everyone, 3))

# Plot all the means and fill in the ones equal to or greater than the actual sample mean

gf_histogram(~ mean, data = SDoM, fill = ~mean>=58)

Example of output from running the code above (your values may vary due to sampling variation):

• Related Articles

• sample distribution

  When sampled data—units that were selected and measured—are shown in such a way where we can see patterns of variation (such as shape, center, and spread); sample distributions can be seen in visualizations such as histograms, boxplots, etc. but also ...

• sampling without replacement

  Sampling without replacement takes a sample from a population, remove the object sampled, then sample again from the remaining objects in the population; the R function sample() does this.

• sampling error

  Sampling error is the variation that occurs from sample to sample due to the fact that no sample is a perfect representation of the population; can be biased or unbiased; also known as sampling variation.

• sampling with replacement

  Sampling with replacement takes a sample from a population, record the values, put all cases back into the population, then sample again; the R function resample() does this.

• distribution triad

  Distribution triad means sample distribution, DGP/population distribution and sampling distribution. ​