Skip to contents

Hand a table to base R

Source: R/tab_classes.R
tabxplor-base-coercion.Rd

as.matrix() gives the table's numbers as a plain numeric matrix; as.table() gives the same matrix as a base table, its dimnames named after the row and column variables. That is the shape base R and the packages built on it expect — a correspondence analysis, chisq.test(), mosaicplot():

FactoMineR::CA(as.matrix(tab(forcats::gss_cat, race, marital)), graph = FALSE)

Only the DATA cells come across. The total row, the total columns and the display-time rows (the base count, add_pct, the p-value and model-fit lines) are dropped, because a test or an analysis run on a table's own margins is wrong; totals = TRUE keeps them. Each cell contributes the number it shows, so a plain tab gives counts, a pct = "row" table proportions, and a numeric column means.

Usage

# S3 method for class 'tabxplor_tab'
as.matrix(x, totals = FALSE, ...)

# S3 method for class 'tabxplor_tab'
as.table(x, totals = FALSE, ...)

Arguments

x

A table made with tab, tab_counts or tab_reg.

totals

Set to TRUE to keep the total row, the total columns and the display-time rows.

...

Not used.

Value

A numeric matrix, or a base table.

Functions

  • as.matrix(tabxplor_tab): the table's numbers as a numeric matrix

  • as.table(tabxplor_tab): the same, as a base table with named dimnames

See also

get_num, tab_export.

Examples

tabs <- tab(forcats::gss_cat, race, marital)
as.matrix(tabs)
#>       No answer Never married Separated Divorced Widowed Married
#> Other         2           633       110      212      70     932
#> Black         2          1305       196      495     262     869
#> White        13          3478       437     2676    1475    8316
as.table(tabs)
#>        marital
#> race    No answer Never married Separated Divorced Widowed Married
#>   Other         2           633       110      212      70     932
#>   Black         2          1305       196      495     262     869
#>   White        13          3478       437     2676    1475    8316

# a row-percentage table gives proportions, not counts:
as.matrix(tab(forcats::gss_cat, race, marital, pct = "row"))
#>          No answer Never married  Separated  Divorced    Widowed   Married
#> Other 0.0010209290     0.3231240 0.05615110 0.1082185 0.03573252 0.4757529
#> Black 0.0006391818     0.4170662 0.06263982 0.1581975 0.08373282 0.2777245
#> White 0.0007929247     0.2121378 0.02665447 0.1632205 0.08996645 0.5072278