sam: Sample from a sampling frame
In gmonette/WWCa: Tools for poststratification

Generate a probability sample from a sampling frame

sam takes a sample size, a 'population' data frame, a formula specifying the count variable and stratification variable(s), an optional expression to select a subset of the population, relative sampling factor and the probability parameter for a binomial response. It generates a sample based on these parameters.

1 2	sam(N = 1, pop = popagetable, fmla = population ~ sex + age + raceethnicity + state, subset, fac, prob)

`N`	sample size (default: 1)
`pop`	a data frame giving frequencies in each stratum defined by a combination of variables. The stratification can be finer than the stratification given in `fmla` (Default: `popagetable`)
`fmla`	a formula of the form `count ~ a + ... + c` where `count` is a variable in `pop` giving the population count in each row of `pop` and the right hand side of the formula identifies the variables used to aggregate `pop` to form a sampling frame (Default: `population ~ sex + age + raceethnicity + state`). The default aggregates over counties in the default value of `pop`
`subset`	an expression that selects a subset, e.g. state == 'AK' & age > '5 to 9 years' (Default: no subsetting)
`fac`	a data frame or a list of data frames whose variables include some subset of variables in the right hand side of `fmla` and a variable named `fac` giving relative sampling factors for each combination of values of the stratification variables. Omitted combinations of variables get a default relative sampling factor of 1. If `fac` is a list, the relative sampling factors from each element are multiplied.
`prob`	a data frame or a list of data frames similar to `fac`, except that a variable named `prob` is used to specify the binomial probability of a '1' in each stratum of respondents. If `prob` is a list, the probabilities from each data frame are combined by adding their logits.