Skip to contents

Get and Set Contrast Matrix Column Names

contr_colnames.Rd

Set and view the column names of contrasts associated with a factor.

Usage

contr_colnames(x)

contr_colnames(x) <- value

contr_colpfx(x) <- value

Arguments

x

a factor for which the contrast column headings are to be set or viewed.

value

a character vector containing the contrast column names, of length equal to the number of contrasts; or in the case of contr_colpfx()<- a character or string used to prefix the existing contrast column names.

Details

contr_colnames() returns the current column names of the contrasts for a factor.

contr_colnames()<- sets the column names of the contrasts for a factor.

cntr_pfx()<- prefixes the current column names of the contrasts for a factor with the character or string provided. This can be useful when factor levels are elided with the factor name as, for instance, in the printed output of summary.glm.

If contrasts are not set for x, both contr_colnames()<- and cntr_pfx()<- set the contrast attribute using the default function from options("contrasts") before modifying the column names.

See also

contrast, contrasts and factor.

Other contrast-names: helm_names()

Examples

(d <- data.frame(
  f = gl(5, 5, labels = LETTERS[1:5]),
  dv = sample(c(0,1), 25, replace = TRUE)
))
#>    f dv
#> 1  A  1
#> 2  A  0
#> 3  A  0
#> 4  A  1
#> 5  A  1
#> 6  B  0
#> 7  B  1
#> 8  B  0
#> 9  B  1
#> 10 B  1
#> 11 C  1
#> 12 C  1
#> 13 C  1
#> 14 C  0
#> 15 C  0
#> 16 D  1
#> 17 D  1
#> 18 D  1
#> 19 D  0
#> 20 D  1
#> 21 E  1
#> 22 E  1
#> 23 E  0
#> 24 E  1
#> 25 E  0

contrasts(d$f) <- contr.helmert
contrasts(d$f)
#>   [,1] [,2] [,3] [,4]
#> A   -1   -1   -1   -1
#> B    1   -1   -1   -1
#> C    0    2   -1   -1
#> D    0    0    3   -1
#> E    0    0    0    4
contr_colnames(d$f)
#> NULL
glm(dv ~ f, family = binomial, data = d) |> summary()
#> 
#> Call:
#> glm(formula = dv ~ f, family = binomial, data = d)
#> 
#> Coefficients:
#>               Estimate Std. Error z value Pr(>|z|)
#> (Intercept)  6.016e-01  4.282e-01   1.405    0.160
#> f1          -2.931e-16  6.455e-01   0.000    1.000
#> f2          -1.672e-16  3.727e-01   0.000    1.000
#> f3           2.452e-01  3.090e-01   0.794    0.427
#> f4          -4.904e-02  2.067e-01  -0.237    0.812
#> 
#> (Dispersion parameter for binomial family taken to be 1)
#> 
#>     Null deviance: 32.671  on 24  degrees of freedom
#> Residual deviance: 31.924  on 20  degrees of freedom
#> AIC: 41.924
#> 
#> Number of Fisher Scoring iterations: 4
#> 

contr_colnames(d$f) <- c("A v. B", "AB v. C", "ABC v. D", "ABCD v. E")
contr_colnames(d$f)
#> [1] "A v. B"    "AB v. C"   "ABC v. D"  "ABCD v. E"
contrasts(d$f)
#>   A v. B AB v. C ABC v. D ABCD v. E
#> A     -1      -1       -1        -1
#> B      1      -1       -1        -1
#> C      0       2       -1        -1
#> D      0       0        3        -1
#> E      0       0        0         4
glm(dv ~ f, family = binomial, data = d) |> summary()
#> 
#> Call:
#> glm(formula = dv ~ f, family = binomial, data = d)
#> 
#> Coefficients:
#>               Estimate Std. Error z value Pr(>|z|)
#> (Intercept)  6.016e-01  4.282e-01   1.405    0.160
#> fA v. B     -2.931e-16  6.455e-01   0.000    1.000
#> fAB v. C    -1.672e-16  3.727e-01   0.000    1.000
#> fABC v. D    2.452e-01  3.090e-01   0.794    0.427
#> fABCD v. E  -4.904e-02  2.067e-01  -0.237    0.812
#> 
#> (Dispersion parameter for binomial family taken to be 1)
#> 
#>     Null deviance: 32.671  on 24  degrees of freedom
#> Residual deviance: 31.924  on 20  degrees of freedom
#> AIC: 41.924
#> 
#> Number of Fisher Scoring iterations: 4
#> 

contr_colpfx(d$f) <- ": "
contr_colnames(d$f)
#> [1] ": A v. B"    ": AB v. C"   ": ABC v. D"  ": ABCD v. E"
glm(dv ~ f, family = binomial, data = d) |> summary()
#> 
#> Call:
#> glm(formula = dv ~ f, family = binomial, data = d)
#> 
#> Coefficients:
#>                Estimate Std. Error z value Pr(>|z|)
#> (Intercept)   6.016e-01  4.282e-01   1.405    0.160
#> f: A v. B    -2.931e-16  6.455e-01   0.000    1.000
#> f: AB v. C   -1.672e-16  3.727e-01   0.000    1.000
#> f: ABC v. D   2.452e-01  3.090e-01   0.794    0.427
#> f: ABCD v. E -4.904e-02  2.067e-01  -0.237    0.812
#> 
#> (Dispersion parameter for binomial family taken to be 1)
#> 
#>     Null deviance: 32.671  on 24  degrees of freedom
#> Residual deviance: 31.924  on 20  degrees of freedom
#> AIC: 41.924
#> 
#> Number of Fisher Scoring iterations: 4
#> 

rm(d)