Calculate Weighted Summaries

Calculate weighted

• means

• medians

• MADs (mean absolute deviations)

• variances

• standard deviations.

These functions all work with ordinary vectors and with rvecs.

Usage

weighted_mean(x, wt = NULL, na_rm = FALSE)

# Default S3 method
weighted_mean(x, wt = NULL, na_rm = FALSE)

# S3 method for class 'rvec'
weighted_mean(x, wt = NULL, na_rm = FALSE)

weighted_mad(x, wt = NULL, na_rm = FALSE)

# Default S3 method
weighted_mad(x, wt = NULL, na_rm = FALSE)

# S3 method for class 'rvec'
weighted_mad(x, wt = NULL, na_rm = FALSE)

weighted_median(x, wt = NULL, na_rm = FALSE)

# Default S3 method
weighted_median(x, wt = NULL, na_rm = FALSE)

# S3 method for class 'rvec'
weighted_median(x, wt = NULL, na_rm = FALSE)

weighted_sd(x, wt = NULL, na_rm = FALSE)

# Default S3 method
weighted_sd(x, wt = NULL, na_rm = FALSE)

# S3 method for class 'rvec'
weighted_sd(x, wt = NULL, na_rm = FALSE)

weighted_var(x, wt = NULL, na_rm = FALSE)

# Default S3 method
weighted_var(x, wt = NULL, na_rm = FALSE)

# S3 method for class 'rvec'
weighted_var(x, wt = NULL, na_rm = FALSE)

Arguments

x

Quantity being summarised. An ordinary vector or an rvec.

wt

Weights. An ordinary vector, an rvec, or NULL (the default.) If NULL, an unweighted summary is returned.

na_rm

Whether to remove NAs in x or wt before calculating. Default is FALSE. See matrixStats::weightedMean() for a description of the algorithm used.

Value

If x or wt or is rvec, then an rvec of length 1. Otherwise, a scalar.

Details

x and wt must have the same length.

Internally the calculations are done by matrixStats functions such as matrixStats::weightedMean() and matrixStats::colWeightedMeans().

See also

• Functions mean(), median(), mad(), var(), sd() for unweighted data all have methods for rvecs

• The original matrixStats weighted summary functions have additional options not implemented in the functions here.

• weighted.mean() is a base R function for weighted data

• For numeric summaries of draws in an rvec, use draws_median(), draws_mean, draws_quantile(), draws_fun().

Examples

## 'x' is rvec, 'wt' is ordinary vector
v <- rvec(list(c(1, 11),
               c(2, 12),
               c(7, 17)))
weights <- c(40, 80, 72)
weighted_mean(v, wt = weights)
#> <rvec_dbl<2>[1]>
#> [1] 3.667,13.67

## 'x' is ordinary vector, 'wt' is rvec
y <- c(1, 2, 3)
w <- rvec(list(c(100, 200),
               c(210, 889),
               c(200, 200)))
weighted_mean(y, wt = w)
#> <rvec_dbl<2>[1]>
#> [1] 2.196,2
weighted_mean(y, wt = w, na_rm = TRUE)
#> <rvec_dbl<2>[1]>
#> [1] 2.196,2