tieders for ordinal models

[larmarange]

cf. #292 and #156

Some examples:

> library(ordinal)
> clm_mod <- clm(rating ~ temp * contact, data = wine)
>   tidy(clm_mod)
                 term   estimate std.error  statistic      p.value coefficient_type
1                 1|2 -1.4112620 0.5453532 -2.5877943 9.659266e-03            alpha
2                 2|3  1.1435537 0.5096555  2.2437776 2.484671e-02            alpha
3                 3|4  3.3770825 0.6381617  5.2918913 1.210578e-07            alpha
4                 4|5  4.9419823 0.7509113  6.5813133 4.663108e-11            alpha
5            tempwarm  2.3211843 0.7009479  3.3114931 9.279953e-04             beta
6          contactyes  1.3474604 0.6603557  2.0405072 4.129984e-02             beta
7 tempwarm:contactyes  0.3595489 0.9238188  0.3891985 6.971293e-01             beta
>   tidy(clm_mod, conf.int = TRUE)
                 term   estimate std.error  statistic      p.value    conf.low conf.high coefficient_type
1                 1|2 -1.4112620 0.5453532 -2.5877943 9.659266e-03          NA        NA            alpha
2                 2|3  1.1435537 0.5096555  2.2437776 2.484671e-02          NA        NA            alpha
3                 3|4  3.3770825 0.6381617  5.2918913 1.210578e-07          NA        NA            alpha
4                 4|5  4.9419823 0.7509113  6.5813133 4.663108e-11          NA        NA            alpha
5          contactyes  1.3474604 0.6603557  2.0405072 4.129984e-02  0.08378091  2.694828             beta
6            tempwarm  2.3211843 0.7009479  3.3114931 9.279953e-04  0.99435182  3.761793             beta
7 tempwarm:contactyes  0.3595489 0.9238188  0.3891985 6.971293e-01 -1.45985126  2.180286             beta
>   tidy(clm_mod, conf.int = TRUE, conf.type = "Wald", exponentiate = TRUE)
                 term    estimate std.error  statistic      p.value    conf.low   conf.high coefficient_type
1                 1|2   0.2438354 0.5453532 -2.5877943 9.659266e-03  0.08373195   0.7100717            alpha
2                 2|3   3.1378996 0.5096555  2.2437776 2.484671e-02  1.15563177   8.5203732            alpha
3                 3|4  29.2852064 0.6381617  5.2918913 1.210578e-07  8.38386057 102.2945581            alpha
4                 4|5 140.0475848 0.7509113  6.5813133 4.663108e-11 32.14391711 610.1722432            alpha
5          contactyes   3.8476417 0.6603557  2.0405072 4.129984e-02  1.05462698  14.0375196             beta
6            tempwarm  10.1877321 0.7009479  3.3114931 9.279953e-04  2.57887058  40.2462558             beta
7 tempwarm:contactyes   1.4326830 0.9238188  0.3891985 6.971293e-01  0.23431175   8.7600407             beta
>   glance(clm_mod)
  edf   logLik      AIC      BIC df.residual
1   7 -86.4162 186.8324 202.7691          65
>   head(augment(clm_mod))
  rating temp contact    .fitted    .se.fit
1      2 cold      no 0.56229641 0.08849613
2      3 cold      no 0.20864908 0.07883377
3      3 cold     yes 0.43467309 0.08370346
4      4 cold     yes 0.08938852 0.04360852
5      4 warm      no 0.19028226 0.07112858
6      4 warm      no 0.19028226 0.07112858
> 
>   clm_mod2 <- clm(rating ~ temp, nominal = ~ contact, data = wine)
>   tidy(clm_mod2)
             term  estimate std.error statistic      p.value coefficient_type
1 1|2.(Intercept) -1.323043 0.5622762 -2.353013 1.862196e-02            alpha
2 2|3.(Intercept)  1.246444 0.4748240  2.625064 8.663257e-03            alpha
3 3|4.(Intercept)  3.550044 0.6560234  5.411459 6.251335e-08            alpha
4 4|5.(Intercept)  4.660247 0.8604027  5.416356 6.082600e-08            alpha
5  1|2.contactyes -1.615059 1.1618005 -1.390135 1.644880e-01            alpha
6  2|3.contactyes -1.511567 0.5906414 -2.559196 1.049144e-02            alpha
7  3|4.contactyes -1.674756 0.6488373 -2.581164 9.846770e-03            alpha
8  4|5.contactyes -1.050623 0.8965109 -1.171902 2.412364e-01            alpha
9        tempwarm  2.519045 0.5350470  4.708081 2.500591e-06             beta
> 
>   clmm_mod <- clmm(rating ~ temp + contact + (1 | judge), data = wine)
>   tidy(clmm_mod)
        term  estimate std.error statistic      p.value coefficient_type
1        1|2 -1.623667 0.6824432 -2.379197 1.735040e-02            alpha
2        2|3  1.513365 0.6037564  2.506582 1.219047e-02            alpha
3        3|4  4.228527 0.8089694  5.227054 1.722324e-07            alpha
4        4|5  6.088773 0.9724566  6.261228 3.819575e-10            alpha
5   tempwarm  3.062997 0.5953779  5.144626 2.680538e-07             beta
6 contactyes  1.834885 0.5125301  3.580053 3.435243e-04             beta
>   glance(clmm_mod)
  edf    logLik      AIC      BIC
1   7 -81.56541 177.1308 193.0675

[alexpghayes]

Nice! Looks like the tests are failing for some reason? I'd add survey to Suggests too.

I'm totally blown away by how many PRs you've been submitting -- much appreciated!

[larmarange]

A data command was missing in the test !!

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.