Tidying methods for multinomial logistic regression models

These methods tidy the coefficients of multinomial logistic regression models generated by multinom of the nnet package.

# S3 method for multinom
tidy(x, conf.int = FALSE, conf.level = 0.95, exponentiate = FALSE, ...)

Arguments

x	A multinom object returned from nnet::multinom().
conf.int	Logical indicating whether or not to include a confidence interval in the tidied output. Defaults to FALSE.
conf.level	The confidence level to use for the confidence interval if conf.int = TRUE. Must be strictly greater than 0 and less than 1. Defaults to 0.95, which corresponds to a 95 percent confidence interval.
exponentiate	Logical indicating whether or not to exponentiate the the coefficient estimates. This is typical for logistic and multinomial regressions, but a bad idea if there is no log or logit link. Defaults to FALSE.
...	Additional arguments. Not used. Needed to match generic signature only. Cautionary note: Misspelled arguments will be absorbed in ..., where they will be ignored. If the misspelled argument has a default value, the default value will be used. For example, if you pass conf.lvel = 0.9, all computation will proceed using conf.level = 0.95. Additionally, if you pass newdata = my_tibble to an augment() method that does not accept a newdata argument, it will use the default value for the data argument.

See also

tidy(), nnet::multinom()

Other multinom tidiers: glance.multinom()

Value

A tibble::tibble() with columns:

conf.high

Upper bound on the confidence interval for the estimate.

conf.low

Lower bound on the confidence interval for the estimate.

estimate

The estimated value of the regression term.

p.value

The two-sided p-value associated with the observed statistic.

statistic

The value of a T-statistic to use in a hypothesis that the regression term is non-zero.

std.error

The standard error of the regression term.

term

The name of the regression term.

y.value

The response level.

Examples

library(nnet)
library(MASS)

example(birthwt)
#> 
#> brthwt> bwt <- with(birthwt, {
#> brthwt+ race <- factor(race, labels = c("white", "black", "other"))
#> brthwt+ ptd <- factor(ptl > 0)
#> brthwt+ ftv <- factor(ftv)
#> brthwt+ levels(ftv)[-(1:2)] <- "2+"
#> brthwt+ data.frame(low = factor(low), age, lwt, race, smoke = (smoke > 0),
#> brthwt+            ptd, ht = (ht > 0), ui = (ui > 0), ftv)
#> brthwt+ })
#> 
#> brthwt> options(contrasts = c("contr.treatment", "contr.poly"))
#> 
#> brthwt> glm(low ~ ., binomial, bwt)
#> 
#> Call:  glm(formula = low ~ ., family = binomial, data = bwt)
#> 
#> Coefficients:
#> (Intercept)          age          lwt    raceblack    raceother    smokeTRUE  
#>     0.82302     -0.03723     -0.01565      1.19241      0.74068      0.75553  
#>     ptdTRUE       htTRUE       uiTRUE         ftv1        ftv2+  
#>     1.34376      1.91317      0.68020     -0.43638      0.17901  
#> 
#> Degrees of Freedom: 188 Total (i.e. Null);  178 Residual
#> Null Deviance:	    234.7 
#> Residual Deviance: 195.5 	AIC: 217.5
bwt.mu <- multinom(low ~ ., bwt)
#> # weights:  12 (11 variable)
#> initial  value 131.004817 
#> iter  10 value 98.029803
#> final  value 97.737759 
#> converged
tidy(bwt.mu)
#> # A tibble: 11 x 6
#>    y.level term        estimate std.error statistic p.value
#>    <chr>   <chr>          <dbl>     <dbl>     <dbl>   <dbl>
#>  1 1       (Intercept)   0.823    1.24        0.661 0.508  
#>  2 1       age          -0.0372   0.0387     -0.962 0.336  
#>  3 1       lwt          -0.0157   0.00708    -2.21  0.0271 
#>  4 1       raceblack     1.19     0.536       2.22  0.0261 
#>  5 1       raceother     0.741    0.462       1.60  0.109  
#>  6 1       smokeTRUE     0.756    0.425       1.78  0.0755 
#>  7 1       ptdTRUE       1.34     0.481       2.80  0.00518
#>  8 1       htTRUE        1.91     0.721       2.65  0.00794
#>  9 1       uiTRUE        0.680    0.464       1.46  0.143  
#> 10 1       ftv1         -0.436    0.479      -0.910 0.363  
#> 11 1       ftv2+         0.179    0.456       0.392 0.695  
glance(bwt.mu)
#> # A tibble: 1 x 4
#>     edf deviance   AIC  nobs
#>   <dbl>    <dbl> <dbl> <int>
#> 1    11     195.  217.   189

#* This model is a truly terrible model
#* but it should show you what the output looks
#* like in a multinomial logistic regression

fit.gear <- multinom(gear ~ mpg + factor(am), data = mtcars)
#> # weights:  12 (6 variable)
#> initial  value 35.155593 
#> iter  10 value 14.156582
#> iter  20 value 14.031881
#> iter  30 value 14.025659
#> iter  40 value 14.021414
#> iter  50 value 14.019824
#> iter  60 value 14.019278
#> iter  70 value 14.018601
#> iter  80 value 14.018282
#> iter  80 value 14.018282
#> iter  90 value 14.017126
#> final  value 14.015374 
#> converged
tidy(fit.gear)
#> # A tibble: 6 x 6
#>   y.level term        estimate std.error statistic  p.value
#>   <chr>   <chr>          <dbl>     <dbl>     <dbl>    <dbl>
#> 1 4       (Intercept)  -11.2       5.32     -2.10  3.60e- 2
#> 2 4       mpg            0.525     0.268     1.96  5.02e- 2
#> 3 4       factor(am)1   11.9      66.9       0.178 8.59e- 1
#> 4 5       (Intercept)  -18.4      67.9      -0.271 7.87e- 1
#> 5 5       mpg            0.366     0.292     1.25  2.10e- 1
#> 6 5       factor(am)1   22.4       2.17     10.3   4.54e-25
glance(fit.gear)
#> # A tibble: 1 x 4
#>     edf deviance   AIC  nobs
#>   <dbl>    <dbl> <dbl> <int>
#> 1     6     28.0  40.0    32