CRAN Package Check Results for Package catspec

Last updated on 2014-04-24 23:47:59.

Flavor Version Tinstall Tcheck Ttotal Status Flags
r-devel-linux-x86_64-debian-clang 0.97 0.56 10.79 11.35 OK
r-devel-linux-x86_64-debian-gcc 0.97 0.54 10.76 11.30 OK
r-devel-linux-x86_64-fedora-clang 0.97 22.57 NOTE
r-devel-linux-x86_64-fedora-gcc 0.97 21.43 NOTE
r-devel-osx-x86_64-clang 0.97 22.67 OK
r-devel-osx-x86_64-gcc 0.97 ERROR
r-devel-windows-ix86+x86_64 0.97 2.00 17.00 19.00 OK
r-patched-linux-x86_64 0.97 0.52 10.86 11.38 OK
r-patched-solaris-sparc 0.97 125.20 OK
r-patched-solaris-x86 0.97 31.60 OK
r-release-linux-ix86 0.97 1.00 24.00 25.00 OK
r-release-linux-x86_64 0.97 0.52 10.71 11.23 OK
r-release-osx-x86_64-mavericks 0.97 ERROR
r-release-windows-ix86+x86_64 0.97 2.00 18.00 20.00 OK
r-oldrel-windows-ix86+x86_64 0.97 3.00 19.00 22.00 OK

Check Details

Version: 0.97
Check: Rd line widths
Result: NOTE
    Rd file 'ctab.Rd':
     \usage lines wider than 90 characters:
     ctab(...,dec.places=NULL,digits=NULL,type=NULL,style=NULL,row.vars=NULL,col.vars=NULL,percentages=NULL,addmargins=NULL)
    
    Rd file 'sqtab.Rd':
     \examples lines wider than 100 characters:
     OccFather<-gl(5,5,labels=c("Upper nonmanual","Lower nonmanual","Upper manual","Lower manual","Farm"))
     glm.q0<-glm(Freq~OccFather+OccSon+mob.qi(OccFather,OccSon,constrained=TRUE),family=poisson(),data=FHtab)
     glm.hrc1<-glm(Freq~OccFather+OccSon+mob.rc1(OccFather,OccSon,equal=TRUE),family=poisson(),data=FHtab)
     m2 <- mlogit(occupation ~ mob.qi(focc,alt)+mob.unif(focc,alt) | education+race,data=pc, reflevel = "farm")
    
    These lines will be truncated in the PDF manual.
Flavors: r-devel-linux-x86_64-fedora-clang, r-devel-linux-x86_64-fedora-gcc

Version: 0.97
Check: package dependencies
Result: NOTE
    Package suggested but not available for checking: ‘mlogit’
Flavors: r-devel-osx-x86_64-gcc, r-release-osx-x86_64-mavericks

Version: 0.97
Check: Rd cross-references
Result: NOTE
    Package unavailable to check Rd xrefs: ‘mlogit’
Flavors: r-devel-osx-x86_64-gcc, r-release-osx-x86_64-mavericks

Version: 0.97
Check: examples
Result: ERROR
    Running examples in ‘catspec-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: sqtab
    > ### Title: sqtab: models for square tables
    > ### Aliases: sqtab check.square fitmacro mob.cp mob.eqmain mob.qi mob.rc1
    > ### mob.symint mob.unif
    > ### Keywords: multivariate
    >
    > ### ** Examples
    >
    > # Examples of loglinear models for square tables,
    > # from Hout, M. (1983). "Mobility Tables". Sage Publication 07-031
    >
    > # Table from page 11 of "Mobility Tables"
    > # Original source: Featherman D.L., R.M. Hauser. (1978) "Opportunity and Change."
    > # New York: Academic, page 49
    >
    > Freq <- c(
    + 1414, 521, 302, 643, 40,
    + 724, 524, 254, 703, 48,
    + 798, 648, 856, 1676, 108,
    + 756, 914, 771, 3325, 237,
    + 409, 357, 441, 1611, 1832)
    >
    > OccFather<-gl(5,5,labels=c("Upper nonmanual","Lower nonmanual","Upper manual","Lower manual","Farm"))
    > OccSon<-gl(5,1,labels=c("Upper nonmanual","Lower nonmanual","Upper manual","Lower manual","Farm"))
    > FHtab <- data.frame(OccFather,OccSon,Freq)
    >
    > xtabs(Freq~OccFather+OccSon,data=FHtab)
     OccSon
    OccFather Upper nonmanual Lower nonmanual Upper manual Lower manual
     Upper nonmanual 1414 521 302 643
     Lower nonmanual 724 524 254 703
     Upper manual 798 648 856 1676
     Lower manual 756 914 771 3325
     Farm 409 357 441 1611
     OccSon
    OccFather Farm
     Upper nonmanual 40
     Lower nonmanual 48
     Upper manual 108
     Lower manual 237
     Farm 1832
    >
    > # independence model
    > indep<-glm(Freq~OccFather+OccSon,family=poisson(),data=FHtab)
    > summary(indep)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon, family = poisson(),
     data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -20.371 -14.730 -2.554 4.015 44.109
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 6.39925 0.02315 276.383 <2e-16 ***
    OccFatherLower nonmanual -0.25932 0.02804 -9.248 <2e-16 ***
    OccFatherUpper manual 0.33598 0.02423 13.865 <2e-16 ***
    OccFatherLower manual 0.72068 0.02256 31.942 <2e-16 ***
    OccFatherFarm 0.46528 0.02361 19.706 <2e-16 ***
    OccSonLower nonmanual -0.32469 0.02411 -13.468 <2e-16 ***
    OccSonUpper manual -0.44653 0.02500 -17.862 <2e-16 ***
    OccSonLower manual 0.66295 0.01922 34.488 <2e-16 ***
    OccSonFarm -0.59366 0.02618 -22.677 <2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.1 on 24 degrees of freedom
    Residual deviance: 6170.1 on 16 degrees of freedom
    AIC: 6390.8
    
    Number of Fisher Scoring iterations: 6
    
    > fitmacro(indep)
    
    
    deviance: 6170.130
    df: 16
    bic: 6011.745
    aic: 6138.130
    Number of parameters: 9
    Number of cases: 19912
    
    
    >
    > wt <- as.numeric(OccFather != OccSon)
    > qi0<-glm(Freq~OccFather+OccSon,weights=wt,family=poisson(),data=FHtab)
    > # A quasi-independence loglinear model, using structural zeros
    > # (page 23 of "Mobility Tables").
    > # 0 1 1 1 1 values of variable "wt"
    > # 1 0 1 1 1
    > # 1 1 0 1 1
    > # 1 1 1 0 1
    > # 1 1 1 1 0
    > qi0<-glm(Freq~OccFather+OccSon,weights=wt,family=poisson(),data=FHtab)
    > summary(qi0)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon, family = poisson(),
     data = FHtab, weights = wt)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -8.181 -3.336 0.000 1.558 13.465
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 5.91935 0.03433 172.444 < 2e-16 ***
    OccFatherLower nonmanual 0.11956 0.03588 3.332 0.000861 ***
    OccFatherUpper manual 0.70710 0.03188 22.179 < 2e-16 ***
    OccFatherLower manual 0.92052 0.03430 26.838 < 2e-16 ***
    OccFatherFarm 0.44093 0.03230 13.653 < 2e-16 ***
    OccSonLower nonmanual -0.07868 0.02847 -2.764 0.005710 **
    OccSonUpper manual -0.26496 0.03137 -8.446 < 2e-16 ***
    OccSonLower manual 0.77953 0.02591 30.081 < 2e-16 ***
    OccSonFarm -1.74558 0.05210 -33.506 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 6123.71 on 19 degrees of freedom
    Residual deviance: 683.34 on 11 degrees of freedom
    AIC: 858.97
    
    Number of Fisher Scoring iterations: 4
    
    > fitmacro(qi0)
    
    
    deviance: 683.342
    df: 11
    bic: 574.452
    aic: 661.342
    Number of parameters: 9
    Number of cases: 19912
    
    
    >
    > # Quasi-independence using a "dummy factor" to create the design
    > # vectors for the diagonal cells (page 23).
    > # 1 0 0 0 0
    > # 0 2 0 0 0
    > # 0 0 3 0 0
    > # 0 0 0 4 0
    > # 0 0 0 0 5
    > glm.qi<-glm(Freq~OccFather+OccSon+mob.qi(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(glm.qi)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.qi(OccFather, OccSon),
     family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -8.181 -3.336 0.000 1.558 13.465
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 5.91935 0.03433 172.444 < 2e-16 ***
    OccFatherLower nonmanual 0.11956 0.03588 3.332 0.000861 ***
    OccFatherUpper manual 0.70710 0.03188 22.179 < 2e-16 ***
    OccFatherLower manual 0.92052 0.03430 26.838 < 2e-16 ***
    OccFatherFarm 0.44093 0.03230 13.653 < 2e-16 ***
    OccSonLower nonmanual -0.07868 0.02847 -2.764 0.005710 **
    OccSonUpper manual -0.26496 0.03137 -8.446 < 2e-16 ***
    OccSonLower manual 0.77953 0.02591 30.081 < 2e-16 ***
    OccSonFarm -1.74558 0.05210 -33.506 < 2e-16 ***
    mob.qi(OccFather, OccSon)2 1.33483 0.04342 30.741 < 2e-16 ***
    mob.qi(OccFather, OccSon)3 0.30126 0.05512 5.465 4.62e-08 ***
    mob.qi(OccFather, OccSon)4 0.39079 0.04695 8.323 < 2e-16 ***
    mob.qi(OccFather, OccSon)5 0.48983 0.03330 14.711 < 2e-16 ***
    mob.qi(OccFather, OccSon)6 2.89847 0.05774 50.197 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.14 on 24 degrees of freedom
    Residual deviance: 683.34 on 11 degrees of freedom
    AIC: 914.05
    
    Number of Fisher Scoring iterations: 4
    
    > fitmacro(glm.qi)
    
    
    deviance: 683.342
    df: 11
    bic: 574.452
    aic: 661.342
    Number of parameters: 14
    Number of cases: 19912
    
    
    >
    > # Quasi-independence without using the functions
    > # Factor labels prevent numeric comparisons, create numeric versions
    > # of the row and column variables
    > OccFather_n <- unclass(OccFather)
    > OccSon_n <- unclass(OccSon)
    > q1 <- ifelse(OccFather_n==OccSon_n & OccSon_n==1,1,0)
    > q2 <- ifelse(OccFather_n==OccSon_n & OccSon_n==2,1,0)
    > q3 <- ifelse(OccFather_n==OccSon_n & OccSon_n==3,1,0)
    > q4 <- ifelse(OccFather_n==OccSon_n & OccSon_n==4,1,0)
    > q5 <- ifelse(OccFather_n==OccSon_n & OccSon_n==5,1,0)
    > glm.qi2<-glm(Freq~OccFather+OccSon+q1+q2+q3+q4+q5,family=poisson(),data=FHtab)
    > summary(glm.qi2)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + q1 + q2 + q3 + q4 +
     q5, family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -8.181 -3.336 0.000 1.558 13.465
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 5.91935 0.03433 172.444 < 2e-16 ***
    OccFatherLower nonmanual 0.11956 0.03588 3.332 0.000861 ***
    OccFatherUpper manual 0.70710 0.03188 22.179 < 2e-16 ***
    OccFatherLower manual 0.92052 0.03430 26.838 < 2e-16 ***
    OccFatherFarm 0.44093 0.03230 13.653 < 2e-16 ***
    OccSonLower nonmanual -0.07868 0.02847 -2.764 0.005710 **
    OccSonUpper manual -0.26496 0.03137 -8.446 < 2e-16 ***
    OccSonLower manual 0.77953 0.02591 30.081 < 2e-16 ***
    OccSonFarm -1.74558 0.05210 -33.506 < 2e-16 ***
    q1 1.33483 0.04342 30.741 < 2e-16 ***
    q2 0.30126 0.05512 5.465 4.62e-08 ***
    q3 0.39079 0.04695 8.323 < 2e-16 ***
    q4 0.48983 0.03330 14.711 < 2e-16 ***
    q5 2.89847 0.05774 50.197 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.14 on 24 degrees of freedom
    Residual deviance: 683.34 on 11 degrees of freedom
    AIC: 914.05
    
    Number of Fisher Scoring iterations: 4
    
    > fitmacro(glm.qi2)
    
    
    deviance: 683.342
    df: 11
    bic: 574.452
    aic: 661.342
    Number of parameters: 14
    Number of cases: 19912
    
    
    >
    > # Quasi-independence constrained (QPM-C, page 31)
    > # Single immobility parameter
    > # 1 0 0 0 0
    > # 0 1 0 0 0
    > # 0 0 1 0 0
    > # 0 0 0 1 0
    > # 0 0 0 0 1
    > glm.q0<-glm(Freq~OccFather+OccSon+mob.qi(OccFather,OccSon,constrained=TRUE),family=poisson(),data=FHtab)
    > # slightly different results than Hout also found in Stata: L2=2567.658, q0=0.964
    > summary(glm.q0)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.qi(OccFather, OccSon,
     constrained = TRUE), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -17.785 -8.712 1.196 4.545 20.115
    
    Coefficients:
     Estimate Std. Error z value
    (Intercept) 6.16634 0.02193 281.135
    OccFatherLower nonmanual -0.18999 0.02866 -6.628
    OccFatherUpper manual 0.45477 0.02500 18.194
    OccFatherLower manual 0.54946 0.02351 23.369
    OccFatherFarm 0.61469 0.02445 25.142
    OccSonLower nonmanual -0.29151 0.02462 -11.840
    OccSonUpper manual -0.54963 0.02574 -21.350
    OccSonLower manual 0.53357 0.02001 26.670
    OccSonFarm -0.74220 0.02705 -27.441
    mob.qi(OccFather, OccSon, constrained = TRUE)2 0.96449 0.01550 62.221
     Pr(>|z|)
    (Intercept) < 2e-16 ***
    OccFatherLower nonmanual 3.39e-11 ***
    OccFatherUpper manual < 2e-16 ***
    OccFatherLower manual < 2e-16 ***
    OccFatherFarm < 2e-16 ***
    OccSonLower nonmanual < 2e-16 ***
    OccSonUpper manual < 2e-16 ***
    OccSonLower manual < 2e-16 ***
    OccSonFarm < 2e-16 ***
    mob.qi(OccFather, OccSon, constrained = TRUE)2 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.1 on 24 degrees of freedom
    Residual deviance: 2567.7 on 15 degrees of freedom
    AIC: 2790.4
    
    Number of Fisher Scoring iterations: 5
    
    > fitmacro(glm.q0)
    
    
    deviance: 2567.658
    df: 15
    bic: 2419.172
    aic: 2537.658
    Number of parameters: 10
    Number of cases: 19912
    
    
    >
    > # Quasi-symmetry using the symmetric cross-classification (page 23)
    > # 0 1 2 3 4 values of variable "sym"
    > # 1 0 5 6 7
    > # 2 5 0 8 9
    > # 3 6 8 0 10
    > # 4 7 9 10 0 */
    > glm.qsym<-
    + glm(Freq~OccFather+OccSon+mob.symint(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(glm.qsym)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.symint(OccFather,
     OccSon), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -2.3247 -0.4558 0.0000 0.6836 2.0935
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 7.25418 0.02659 272.780 <2e-16 ***
    OccFatherLower nonmanual -0.74821 0.04311 -17.355 <2e-16 ***
    OccFatherUpper manual -0.55598 0.04166 -13.345 <2e-16 ***
    OccFatherLower manual -0.56958 0.04100 -13.891 <2e-16 ***
    OccFatherFarm -1.24550 0.05447 -22.868 <2e-16 ***
    OccSonLower nonmanual -0.89828 0.04450 -20.184 <2e-16 ***
    OccSonUpper manual -1.58756 0.05056 -31.397 <2e-16 ***
    OccSonLower manual -0.85890 0.04341 -19.786 <2e-16 ***
    OccSonFarm -3.51518 0.07705 -45.624 <2e-16 ***
    mob.symint(OccFather, OccSon)2:2 0.32690 0.03820 8.557 <2e-16 ***
    mob.symint(OccFather, OccSon)2:3 0.65816 0.06032 10.910 <2e-16 ***
    mob.symint(OccFather, OccSon)2:4 0.97607 0.05347 18.256 <2e-16 ***
    mob.symint(OccFather, OccSon)2:5 0.78009 0.08074 9.662 <2e-16 ***
    mob.symint(OccFather, OccSon)3:3 0.82082 0.03841 21.370 <2e-16 ***
    mob.symint(OccFather, OccSon)3:4 1.57396 0.05351 29.414 <2e-16 ***
    mob.symint(OccFather, OccSon)3:5 1.63239 0.07953 20.526 <2e-16 ***
    mob.symint(OccFather, OccSon)4:4 1.14176 0.03121 36.583 <2e-16 ***
    mob.symint(OccFather, OccSon)4:5 2.24279 0.06714 33.406 <2e-16 ***
    mob.symint(OccFather, OccSon)5:5 2.50983 0.05607 44.763 <2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.143 on 24 degrees of freedom
    Residual deviance: 27.448 on 6 degrees of freedom
    AIC: 268.16
    
    Number of Fisher Scoring iterations: 3
    
    > fitmacro(glm.qsym)
    
    
    deviance: 27.448
    df: 6
    bic: -31.947
    aic: 15.448
    Number of parameters: 19
    Number of cases: 19912
    
    
    >
    > symmetry<-glm(Freq~mob.eqmain(OccFather,OccSon)
    + +mob.symint(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(symmetry)
    
    Call:
    glm(formula = Freq ~ mob.eqmain(OccFather, OccSon) + mob.symint(OccFather,
     OccSon), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -27.001 -10.116 0.000 8.701 20.423
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 7.25418 0.02659 272.780 <2e-16 ***
    mob.eqmain(OccFather, OccSon)2 -0.82043 0.03886 -21.110 <2e-16 ***
    mob.eqmain(OccFather, OccSon)3 -0.94426 0.04020 -23.487 <2e-16 ***
    mob.eqmain(OccFather, OccSon)4 -0.70381 0.03771 -18.664 <2e-16 ***
    mob.eqmain(OccFather, OccSon)5 -1.84030 0.05417 -33.973 <2e-16 ***
    mob.symint(OccFather, OccSon)2:2 0.32409 0.03817 8.490 <2e-16 ***
    mob.symint(OccFather, OccSon)2:3 0.62198 0.05940 10.471 <2e-16 ***
    mob.symint(OccFather, OccSon)2:4 0.96525 0.05333 18.101 <2e-16 ***
    mob.symint(OccFather, OccSon)2:5 0.71730 0.07878 9.105 <2e-16 ***
    mob.symint(OccFather, OccSon)3:3 0.69331 0.03712 18.677 <2e-16 ***
    mob.symint(OccFather, OccSon)3:4 1.50336 0.05234 28.722 <2e-16 ***
    mob.symint(OccFather, OccSon)3:5 1.14534 0.07527 15.217 <2e-16 ***
    mob.symint(OccFather, OccSon)4:4 1.13134 0.03109 36.385 <2e-16 ***
    mob.symint(OccFather, OccSon)4:5 2.11865 0.06473 32.729 <2e-16 ***
    mob.symint(OccFather, OccSon)5:5 1.96979 0.05040 39.081 <2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.1 on 24 degrees of freedom
    Residual deviance: 2804.9 on 10 degrees of freedom
    AIC: 3037.6
    
    Number of Fisher Scoring iterations: 5
    
    > fitmacro(symmetry)
    
    
    deviance: 2804.873
    df: 10
    bic: 2705.882
    aic: 2784.873
    Number of parameters: 15
    Number of cases: 19912
    
    
    >
    > # Crossings parameter model (page 35)
    > # 0 v1 v1 v1 v1 | 0 0 v2 v2 v2 | 0 0 0 v3 v3 | 0 0 0 0 v4
    > # v1 0 0 0 0 | 0 0 v2 v2 v2 | 0 0 0 v3 v3 | 0 0 0 0 v4
    > # v1 0 0 0 0 | v2 v2 0 0 0 | 0 0 0 v3 v3 | 0 0 0 0 v4
    > # v1 0 0 0 0 | v2 v2 0 0 0 | v3 v3 v3 0 0 | 0 0 0 0 v4
    > # v1 0 0 0 0 | v2 v2 0 0 0 | v3 v3 v3 0 0 | v4 v4 v4 v4 0
    > glm.cp<-glm(Freq~OccFather+OccSon+mob.cp(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(glm.cp)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.cp(OccFather, OccSon),
     family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -3.1888 -0.8059 0.0000 0.6484 5.2990
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 7.25418 0.02659 272.780 < 2e-16 ***
    OccFatherLower nonmanual -0.35986 0.02879 -12.500 < 2e-16 ***
    OccFatherUpper manual 0.20073 0.02568 7.816 5.45e-15 ***
    OccFatherLower manual 0.57421 0.02441 23.521 < 2e-16 ***
    OccFatherFarm 1.25914 0.03463 36.365 < 2e-16 ***
    OccSonLower nonmanual -0.53016 0.02700 -19.637 < 2e-16 ***
    OccSonUpper manual -0.79953 0.02810 -28.449 < 2e-16 ***
    OccSonLower manual 0.28056 0.02344 11.969 < 2e-16 ***
    OccSonFarm -1.00015 0.03463 -28.885 < 2e-16 ***
    mob.cp(OccFather, OccSon)1 -0.42562 0.02394 -17.776 < 2e-16 ***
    mob.cp(OccFather, OccSon)2 -0.36748 0.02115 -17.372 < 2e-16 ***
    mob.cp(OccFather, OccSon)3 -0.29347 0.01728 -16.980 < 2e-16 ***
    mob.cp(OccFather, OccSon)4 -1.40259 0.02894 -48.458 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.143 on 24 degrees of freedom
    Residual deviance: 89.914 on 12 degrees of freedom
    AIC: 318.63
    
    Number of Fisher Scoring iterations: 4
    
    > fitmacro(glm.cp)
    
    
    deviance: 89.914
    df: 12
    bic: -28.875
    aic: 65.914
    Number of parameters: 13
    Number of cases: 19912
    
    
    >
    > # Uniform association model: linear by linear association (page 58)
    > glm.unif<-glm(Freq~OccFather+OccSon+mob.unif(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(glm.unif)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.unif(OccFather,
     OccSon), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -24.103 -3.302 -1.461 4.088 25.781
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 6.980009 0.021329 327.25 <2e-16 ***
    OccFatherLower nonmanual -0.860564 0.029168 -29.50 <2e-16 ***
    OccFatherUpper manual -0.991014 0.031298 -31.66 <2e-16 ***
    OccFatherLower manual -1.457724 0.042152 -34.58 <2e-16 ***
    OccFatherFarm -2.672356 0.059456 -44.95 <2e-16 ***
    OccSonLower nonmanual -1.046149 0.026393 -39.64 <2e-16 ***
    OccSonUpper manual -2.018346 0.035704 -56.53 <2e-16 ***
    OccSonLower manual -1.876387 0.047874 -39.19 <2e-16 ***
    OccSonFarm -4.195326 0.069883 -60.03 <2e-16 ***
    mob.unif(OccFather, OccSon) 0.269259 0.004804 56.05 <2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.1 on 24 degrees of freedom
    Residual deviance: 2280.7 on 15 degrees of freedom
    AIC: 2503.4
    
    Number of Fisher Scoring iterations: 5
    
    > fitmacro(glm.unif)
    
    
    deviance: 2280.694
    df: 15
    bic: 2132.207
    aic: 2250.694
    Number of parameters: 10
    Number of cases: 19912
    
    
    >
    > # RC model 1 (unequal row and column effects, page 58)
    > # Fits a uniform association parameter and row and column effect
    > # parameters. Row and column effect parameters have the
    > # restriction that the first and last categories are zero.
    > glm.rc1<-glm(Freq~OccFather+OccSon+mob.rc1(OccFather,OccSon),family=poisson(),data=FHtab)
    > summary(glm.rc1)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.rc1(OccFather,
     OccSon), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -15.254 -2.454 -1.352 4.013 13.785
    
    Coefficients:
     Estimate Std. Error z value Pr(>|z|)
    (Intercept) 6.672678 0.026313 253.585 < 2e-16 ***
    OccFatherLower nonmanual -1.154320 0.060824 -18.978 < 2e-16 ***
    OccFatherUpper manual -1.617053 0.061608 -26.247 < 2e-16 ***
    OccFatherLower manual -2.048505 0.067952 -30.147 < 2e-16 ***
    OccFatherFarm -3.636547 0.078946 -46.064 < 2e-16 ***
    OccSonLower nonmanual -1.125905 0.057383 -19.621 < 2e-16 ***
    OccSonUpper manual -1.814371 0.066842 -27.144 < 2e-16 ***
    OccSonLower manual -1.376182 0.060296 -22.824 < 2e-16 ***
    OccSonFarm -8.992250 0.214486 -41.925 < 2e-16 ***
    mob.rc1(OccFather, OccSon)R2 0.029654 0.020919 1.418 0.156
    mob.rc1(OccFather, OccSon)R3 0.079253 0.017742 4.467 7.93e-06 ***
    mob.rc1(OccFather, OccSon)R4 0.002588 0.017053 0.152 0.879
    mob.rc1(OccFather, OccSon)U 0.533707 0.011078 48.178 < 2e-16 ***
    mob.rc1(OccFather, OccSon)C2 -0.252840 0.020567 -12.293 < 2e-16 ***
    mob.rc1(OccFather, OccSon)C3 -0.614921 0.028553 -21.536 < 2e-16 ***
    mob.rc1(OccFather, OccSon)C4 -0.960196 0.034636 -27.722 < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.14 on 24 degrees of freedom
    Residual deviance: 877.81 on 9 degrees of freedom
    AIC: 1112.5
    
    Number of Fisher Scoring iterations: 5
    
    > fitmacro(glm.rc1)
    
    
    deviance: 877.808
    df: 9
    bic: 788.717
    aic: 859.808
    Number of parameters: 16
    Number of cases: 19912
    
    
    >
    > # Homogeneous row and column effects model 1 (page 58)
    > # An equality restriction is placed on the row and column effects
    > glm.hrc1<-glm(Freq~OccFather+OccSon+mob.rc1(OccFather,OccSon,equal=TRUE),family=poisson(),data=FHtab)
    > # Results differ from those in Hout, replicated by other programs
    > summary(glm.hrc1)
    
    Call:
    glm(formula = Freq ~ OccFather + OccSon + mob.rc1(OccFather,
     OccSon, equal = TRUE), family = poisson(), data = FHtab)
    
    Deviance Residuals:
     Min 1Q Median 3Q Max
    -15.800 -3.195 -1.188 3.070 17.434
    
    Coefficients:
     Estimate Std. Error z value
    (Intercept) 6.798987 0.025260 269.157
    OccFatherLower nonmanual -0.982508 0.047712 -20.592
    OccFatherUpper manual -1.022509 0.048903 -20.909
    OccFatherLower manual -1.141834 0.050288 -22.706
    OccFatherFarm -3.752391 0.083624 -44.872
    OccSonLower nonmanual -1.170357 0.050894 -22.996
    OccSonUpper manual -2.049682 0.057562 -35.608
    OccSonLower manual -1.570274 0.055556 -28.264
    OccSonFarm -5.599423 0.100750 -55.577
    mob.rc1(OccFather, OccSon, equal = TRUE)RC2 -0.035255 0.012639 -2.789
    mob.rc1(OccFather, OccSon, equal = TRUE)RC3 -0.141634 0.011321 -12.510
    mob.rc1(OccFather, OccSon, equal = TRUE)RC4 -0.313500 0.011305 -27.730
    mob.rc1(OccFather, OccSon, equal = TRUE)U 0.392262 0.007312 53.649
     Pr(>|z|)
    (Intercept) < 2e-16 ***
    OccFatherLower nonmanual < 2e-16 ***
    OccFatherUpper manual < 2e-16 ***
    OccFatherLower manual < 2e-16 ***
    OccFatherFarm < 2e-16 ***
    OccSonLower nonmanual < 2e-16 ***
    OccSonUpper manual < 2e-16 ***
    OccSonLower manual < 2e-16 ***
    OccSonFarm < 2e-16 ***
    mob.rc1(OccFather, OccSon, equal = TRUE)RC2 0.00528 **
    mob.rc1(OccFather, OccSon, equal = TRUE)RC3 < 2e-16 ***
    mob.rc1(OccFather, OccSon, equal = TRUE)RC4 < 2e-16 ***
    mob.rc1(OccFather, OccSon, equal = TRUE)U < 2e-16 ***
    ---
    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    
    (Dispersion parameter for poisson family taken to be 1)
    
     Null deviance: 13132.1 on 24 degrees of freedom
    Residual deviance: 1451.3 on 12 degrees of freedom
    AIC: 1680.1
    
    Number of Fisher Scoring iterations: 5
    
    > fitmacro(glm.hrc1)
    
    
    deviance: 1451.347
    df: 12
    bic: 1332.558
    aic: 1427.347
    Number of parameters: 13
    Number of cases: 19912
    
    
    >
    > #-------------------------------------------------------------------------------
    > # Examples on using these models in multinomial logistic regression
    > #-------------------------------------------------------------------------------
    > # Data from the 1972-78 GSS used by Logan (1983)
    > library(survival)
    Loading required package: splines
    > data(logan)
    >
    > # Restructure the data in 'long' format using mlogit.data
    > library(mlogit)
    Error in library(mlogit) : there is no package called ‘mlogit’
    Execution halted
Flavors: r-devel-osx-x86_64-gcc, r-release-osx-x86_64-mavericks