R Under development (unstable) (2025-08-24 r88696 ucrt) -- "Unsuffered Consequences"
Copyright (C) 2025 The R Foundation for Statistical Computing
Platform: x86_64-w64-mingw32/x64

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> ## Clogg & Shihadeh (1994), Tables 3.5a and b (p. 55-61)
> # See also ?rc
> 
> library(logmult)
Loading required package: gnm

Attaching package: 'logmult'

The following object is masked from 'package:gnm':

    se

> data(gss88)
> model <- rc(gss88, start=NA)
Running base model to find starting values...
Running real model...
Initialising
Running main iterations..............................
Done
> 
> # Unweighted scores
> summary(model, weighting="none")
Call:
rc(tab = gss88, start = NA)

Deviance Residuals:
    Min       1Q   Median       3Q      Max  
-2.7608  -0.4975   0.0749   0.4650   1.9374  

Association coefficients:
                                Normalized  Adjusted
Dim1                              8.075599        NA
Dim1:OccupationProfessional-I     0.565514    1.6071
Dim1:OccupationProfessional-II    0.564018    1.6028
Dim1:OccupationManagers           0.146153    0.4153
Dim1:OccupationClerical          -0.013199   -0.0375
Dim1:OccupationCraftsmen-I       -0.232452   -0.6606
Dim1:OccupationCraftsmen-II       0.037196    0.1057
Dim1:OccupationOperatives (NT)   -0.417400   -1.1861
Dim1:OccupationOperatives        -0.196173   -0.5575
Dim1:OccupationFarmers           -0.189421   -0.5383
Dim1:OccupationService           -0.160650   -0.4565
Dim1:OccupationMissing           -0.103587   -0.2944
Dim1:Years.of.Schooling0-7       -0.561801   -1.5965
Dim1:Years.of.Schooling8-11      -0.409544   -1.1638
Dim1:Years.of.Schooling12        -0.084106   -0.2390
Dim1:Years.of.Schooling13-15      0.065232    0.1854
Dim1:Years.of.Schooling16         0.408340    1.1604
Dim1:Years.of.Schooling17+        0.581879    1.6536

No standard errors available
(jackknife and bootstrap disabled, or weighting changed since fitting).

Normalization weights: none
Deviance:              53.47085
Pearson chi-squared:   52.12034
Dissimilarity index:   7.179518%
Residual df:           36
BIC:                   -200.8605
AIC:                   -18.52915
> # Marginally weighted scores
> summary(model, weighting="marginal")
Call:
rc(tab = gss88, start = NA)

Deviance Residuals:
    Min       1Q   Median       3Q      Max  
-2.7608  -0.4975   0.0749   0.4650   1.9374  

Association coefficients:
                                Normalized  Adjusted
Dim1                             0.7320591        NA
Dim1:OccupationProfessional-I    1.7834155    1.5259
Dim1:OccupationProfessional-II   1.7783737    1.5216
Dim1:OccupationManagers          0.3700858    0.3166
Dim1:OccupationClerical         -0.1669616   -0.1429
Dim1:OccupationCraftsmen-I      -0.9058900   -0.7751
Dim1:OccupationCraftsmen-II      0.0028784    0.0025
Dim1:OccupationOperatives (NT)  -1.5291990   -1.3084
Dim1:OccupationOperatives       -0.7836197   -0.6705
Dim1:OccupationFarmers          -0.7608640   -0.6510
Dim1:OccupationService          -0.6639001   -0.5680
Dim1:OccupationMissing          -0.4715891   -0.4035
Dim1:Years.of.Schooling0-7      -1.8890420   -1.6163
Dim1:Years.of.Schooling8-11     -1.3906757   -1.1899
Dim1:Years.of.Schooling12       -0.3254488   -0.2785
Dim1:Years.of.Schooling13-15     0.1633656    0.1398
Dim1:Years.of.Schooling16        1.2864265    1.1007
Dim1:Years.of.Schooling17+       1.8544550    1.5867

No standard errors available
(jackknife and bootstrap disabled, or weighting changed since fitting).

Normalization weights: marginal
Deviance:              53.47085
Pearson chi-squared:   52.12034
Dissimilarity index:   7.179518%
Residual df:           36
BIC:                   -200.8605
AIC:                   -18.52915
> # Uniformly weighted scores
> summary(model, weighting="uniform")
Call:
rc(tab = gss88, start = NA)

Deviance Residuals:
    Min       1Q   Median       3Q      Max  
-2.7608  -0.4975   0.0749   0.4650   1.9374  

Association coefficients:
                                Normalized  Adjusted
Dim1                              0.994038        NA
Dim1:OccupationProfessional-I     1.875598    1.8700
Dim1:OccupationProfessional-II    1.870636    1.8651
Dim1:OccupationManagers           0.484735    0.4833
Dim1:OccupationClerical          -0.043775   -0.0436
Dim1:OccupationCraftsmen-I       -0.770957   -0.7687
Dim1:OccupationCraftsmen-II       0.123365    0.1230
Dim1:OccupationOperatives (NT)   -1.384358   -1.3802
Dim1:OccupationOperatives        -0.650631   -0.6487
Dim1:OccupationFarmers           -0.628237   -0.6264
Dim1:OccupationService           -0.532814   -0.5312
Dim1:OccupationMissing           -0.343560   -0.3425
Dim1:Years.of.Schooling0-7       -1.376125   -1.3720
Dim1:Years.of.Schooling8-11      -1.003175   -1.0002
Dim1:Years.of.Schooling12        -0.206016   -0.2054
Dim1:Years.of.Schooling13-15      0.159786    0.1593
Dim1:Years.of.Schooling16         1.000224    0.9972
Dim1:Years.of.Schooling17+        1.425306    1.4211

No standard errors available
(jackknife and bootstrap disabled, or weighting changed since fitting).

Normalization weights: uniform
Deviance:              53.47085
Pearson chi-squared:   52.12034
Dissimilarity index:   7.179518%
Residual df:           36
BIC:                   -200.8605
AIC:                   -18.52915
> 
> unweighted <- assoc(model, weighting="none")
> marginal <- assoc(model, weighting="marginal")
> uniform <- assoc(model, weighting="uniform")
> 
> stopifnot(all.equal(round(c(unweighted$row, unweighted$row * sqrt(unweighted$phi[1]),
+                             marginal$row, marginal$row * sqrt(marginal$phi[1]),
+                             uniform$row, uniform$row * sqrt(uniform$phi[1])), 2),
+                     c(0.57, 0.56, 0.15, -0.01, -0.23, 0.04, -0.42, -0.2, -0.19, -0.16, -0.1,
+                       1.61, 1.6, 0.42, -0.04, -0.66, 0.11, -1.19, -0.56, -0.54, -0.46, -0.29,
+                       1.78, 1.78, 0.37, -0.17, -0.91, 0, -1.53, -0.78, -0.76, -0.66, -0.47,
+                       1.53, 1.52, 0.32, -0.14, -0.78, 0, -1.31, -0.67, -0.65, -0.57, -0.4,
+                       1.88, 1.87, 0.48, -0.04,  -0.77, 0.12, -1.38, -0.65, -0.63, -0.53, -0.34, 
+                       1.87, 1.87, 0.48, -0.04, -0.77, 0.12, -1.38, -0.65, -0.63, -0.53, -0.34)))
> stopifnot(all.equal(round(c(unweighted$col, unweighted$col * sqrt(unweighted$phi[1]),
+                             marginal$col, marginal$col * sqrt(marginal$phi[1]),
+                             uniform$col, uniform$col * sqrt(uniform$phi[1])), 2),
+                     c(-0.56, -0.41, -0.08, 0.07, 0.41, 0.58,
+                       -1.6, -1.16, -0.24, 0.19, 1.16, 1.65,
+                       -1.89, -1.39, -0.33, 0.16, 1.29, 1.85,
+                       -1.62, -1.19, -0.28, 0.14, 1.1, 1.59,
+                       -1.38, -1, -0.21, 0.16, 1, 1.43,
+                       -1.37, -1, -0.21, 0.16, 1, 1.42)))
> 
> # Test anova
> indep <- gnm(Freq ~ Occupation + Years.of.Schooling, data=gss88, family=poisson)
> anova(indep, unweighted, test="LR")
Analysis of Deviance Table

Model: poisson, link: log

Response: Freq

Terms added sequentially (first to last)


                   Df Deviance Resid. Df Resid. Dev  Pr(>Chi)    
NULL                                  65    1359.84              
Occupation         10   464.10        55     895.75 < 2.2e-16 ***
Years.of.Schooling  5   440.69        50     455.06 < 2.2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
> 
> proc.time()
   user  system elapsed 
   1.26    0.17    1.39