Package index • drawn

Drawing a sample

One verb, and the designs it takes.

draw(): Draw a sample
designs: Sampling designs
plot(<drawn_design>): Picture a sampling design
is_design(): Is this a sampling design?
draw_design(): Design-specific draw method

What a design knows before you draw

The probabilities that make a sample estimable.

inclusion_prob() sampling_weight(): Inclusion probabilities and design weights
joint_prob(): Joint inclusion probabilities

Estimating from a sample

Totals, means and proportions, overall or by domain, with the variance estimator the design calls for and a design effect.

ht_total(): Estimate a population total from a sample
ht_mean(): Estimate a population mean
deff(): Design effect

Planning and checking

How many rows to draw, what you actually got, and where to go next.

plan_size(): How large a sample do you need?
sample_summary(): Describe a drawn sample
as_svydesign(): Hand a sample to the survey package

Equal-probability designs

Every unit equally likely, or equally likely within a group.

design_simple(): Simple random sampling
design_systematic(): Systematic sampling
design_stratified(): Stratified sampling
design_reservoir(): Reservoir sampling

Cluster designs

Sample groups, then optionally sample within them.

design_cluster(): Cluster sampling
design_multistage(): Multi-stage sampling

Unequal-probability designs

Selection driven by size, and units too big to leave to chance.

design_weighted(): Weighted sampling
design_certainty(): Take some rows with certainty, sample the rest

Space and time

A sample spread evenly over a map or over covariates, restricted to a region, or taken across time intervals.

design_spread(): Spatially balanced sampling
design_spatial(): Sampling within a region
design_temporal(): Temporal sampling

Resampling

design_bootstrap(): Bootstrap sampling