getHypothesisTest {mobilityIndexR}

Hypothesis Test for Two Mobility Datasets

Description

Calculates hypothesis tests of mobility indices from two datasets. Specifically, for datasets A and B, this function performs one-sided nonparametric hypothesis tests that the index value for A is greater than the corresponding index value for B. Supports Prais-Bibby, Absolute Movement, Origin Specific, and Weighted Group Mobility indices and relative, mixed, and absolute types of rankings in the calculation these indices.

Usage

getHypothesisTest(
  dat_A,
  dat_B,
  cols_A,
  cols_B,
  type,
  indices = "all",
  num_ranks,
  exclude_value,
  bounds,
  rerank_exclude_value = FALSE,
  strict = TRUE,
  bootstrap_iter = 100
)

Arguments

dat_A

a dataframe with an "id" column

dat_B

a dataframe with an "id" column

cols_A

a list of character strings denoting the first and second column to be used in the index calculations for dat_A

cols_B

a list of character strings denoting the first and second column to be used in the index calculations for dat_B

type

a character string indicating the type of ranking; accepts 'relative', 'mixed', and 'absolute'

indices

a vector of character strings indicating which mobility indices are desired; currently support 'prais_bibby', 'average_movement', 'wgm', and 'origin_specific'. The default value is 'all'.

num_ranks

an integer specifying the number of ranks for a relative or mixed ranking

exclude_value

a single numeric value that is excluded in calculating the transition matrix; see the rerank_exclude_value parameter to specify how the exclude value is handled

bounds

a sequence of numeric bounds for defining absolute ranks

rerank_exclude_value

a character string indicating how the exclude value is handled when present; accepts 'as_new_rank', 'as_existing_rank', and 'exclude'

strict

logical. If TRUE, indices are calculated from the given values. If FALSE, indices are calculated by jittering the values to ensure uniqueness of bounds of ranks. Only used with relative and mixed types. The default value is TRUE.

bootstrap_iter

the number of bootstrap iterations used to estimate hypothesis tests. The default value is 100.

Value

Returns a named vector containing the estimated probabilities that index value for dataset A is greater than the corresponding index value for dataset B

Examples

getHypothesisTest(dat_A = incomeMobility,
                  dat_B = incomeMobility,
                  cols_A = c("t0", "t3"),
                  cols_B = c("t5", "t8"),
                  type = "relative",
                  num_ranks = 5)