Calculate Conditional Fractions from a Contingency Table

Converts contingency-table counts into conditional probabilities (fractions).

Usage

calculate_fractions(dt, group_by, target_attribute, margins_group = NULL)

Arguments

dt

A contingency table containing a count column.

group_by

Character vector containing the conditioning variables.

target_attribute

Character string identifying the target attribute.

margins_group

Optional character vector containing additional grouping variables introduced through margin fitting.

Value

The supplied contingency table with an additional column named fraction.

Details

For each conditioning group, the function computes the proportion represented by each target-attribute category.

The resulting fractions are subsequently used to allocate synthetic agents during conditional attribute assignment.

Fractions are calculated separately within each conditioning group.

For a contingency table:

age_group gender education count
18-64     Male   Degree    45
18-64     Male   Diploma   25
18-64     Male   School    30

the resulting fractions are:

Degree   0.45
Diploma  0.25
School   0.30

Groups whose total count equals zero receive fractions of zero rather than NA or NaN.

This behaviour prevents failures during synthetic-population generation when contingency tables contain zero-count groups.

See also

getGroupFractions, calculate_group_counts, ReplicaAdder, enhance

Examples

library(data.table)
#> 
#> Attaching package: ‘data.table’
#> The following object is masked from ‘package:base’:
#> 
#>     %notin%

dt <- data.table(
  age_group = c(
    "18-64",
    "18-64",
    "18-64"
  ),
  gender = c(
    "Male",
    "Male",
    "Male"
  ),
  education = c(
    "Degree",
    "Diploma",
    "School"
  ),
  count = c(
    45,
    25,
    30
  )
)

calculate_fractions(
  dt,
  group_by = c(
    "age_group",
    "gender"
  ),
  target_attribute =
    "education"
)
#>    age_group gender education count fraction
#>       <char> <char>    <char> <num>    <num>
#> 1:     18-64   Male    Degree    45     0.45
#> 2:     18-64   Male   Diploma    25     0.25
#> 3:     18-64   Male    School    30     0.30

dt <- data.table(
  gender = c(
    "Female",
    "Female",
    "Female"
  ),
  education = c(
    "Degree",
    "Diploma",
    "School"
  ),
  count = c(
    0,
    0,
    0
  )
)

calculate_fractions(
  dt,
  group_by = "gender",
  target_attribute = "education"
)
#>    gender education count fraction
#>    <char>    <char> <num>    <num>
#> 1: Female    Degree     0        0
#> 2: Female   Diploma     0        0
#> 3: Female    School     0        0