R/lms.R
summary_c2c.Rdadjusting lm object results according to original number of degree of freedom. The standard errors, t statistics and p values have to be adjusted because of replicated observations.
summary_c2c(x, df_old, df_new = x$df.residual)lm object
integer number of d.f in original dataset. For bigger datasets `nrow` should be sufficient.
integer number of d.f in dataset with replicated rows, Default: x$df.residual
data.frame with additional columns over a regular summary.lm output, like correct and statistics adjusted by it.
The size of the correction is equal to sqrt(df_new / df_old).
Where standard errors are multiplied and t statistics divided by it.
In most cases the default df_new value should be used.
data("occup_small", package = "cat2cat")
data("trans", package = "cat2cat")
occup_old <- occup_small[occup_small$year == 2008, ]
occup_new <- occup_small[occup_small$year == 2010, ]
occup_2 <- cat2cat(
data = list(
old = occup_old,
new = occup_new,
cat_var = "code",
time_var = "year"
),
mappings = list(trans = trans, direction = "backward"),
ml = list(
data = occup_new,
cat_var = "code",
method = "knn",
features = c("age", "sex", "edu", "exp", "parttime", "salary"),
args = list(k = 10)
)
)
# Regression
# we have to adjust size of std as we artificialy enlarge degrees of freedom
lms <- lm(
formula = I(log(salary)) ~ age + sex + factor(edu) + parttime + exp,
data = occup_2$old,
weights = multiplier * wei_freq_c2c
)
summary_c2c(lms, df_old = nrow(occup_old))
#> Estimate Std. Error t value Pr(>|t|) correct
#> (Intercept) 8.642709060 0.0305125691 283.250782 0.000000e+00 2.412389
#> age -0.002354694 0.0007645294 -3.079926 2.075450e-03 2.412389
#> sexTRUE 0.291655854 0.0086207208 33.831957 3.607064e-239 2.412389
#> factor(edu)2 -0.063704597 0.0161174380 -3.952526 7.779765e-05 2.412389
#> factor(edu)3 -0.338008780 0.0201287688 -16.792323 1.565695e-62 2.412389
#> factor(edu)4 -0.381827051 0.0123532071 -30.909143 1.641476e-201 2.412389
#> factor(edu)5 -0.366016945 0.0170324446 -21.489396 1.904297e-100 2.412389
#> factor(edu)6 -0.579703259 0.0123940057 -46.772873 0.000000e+00 2.412389
#> factor(edu)7 -0.523761245 0.0890371147 -5.882505 4.154668e-09 2.412389
#> factor(edu)8 -0.630390830 0.0184889170 -34.095606 1.054062e-242 2.412389
#> parttime 1.875598705 0.0200106821 93.729874 0.000000e+00 2.412389
#> exp 0.010307126 0.0007306681 14.106440 8.256218e-45 2.412389
#> std.error_c statistic_c p.value_c
#> (Intercept) 0.073608193 117.415041 0.000000e+00
#> age 0.001844343 -1.276712 2.018554e-01
#> sexTRUE 0.020796534 14.024253 1.195445e-42
#> factor(edu)2 0.038881534 -1.638428 1.014932e-01
#> factor(edu)3 0.048558425 -6.960868 4.597874e-12
#> factor(edu)4 0.029800744 -12.812668 3.782334e-36
#> factor(edu)5 0.041088886 -8.907931 1.160653e-18
#> factor(edu)6 0.029899166 -19.388610 9.622762e-77
#> factor(edu)7 0.214792178 -2.438456 1.483856e-02
#> factor(edu)8 0.044602464 -14.133543 2.935764e-43
#> parttime 0.048273554 38.853545 2.899983e-245
#> exp 0.001762656 5.847498 5.833868e-09