Table of Regression Results

Extract a table of results for each response

Usage

results(fit, coef = NULL, ddf = c("finite", "asymptotic"), ...)

# S4 method for class 'modelFitsResponses'
results(fit, coef, ddf = c("finite", "asymptotic"), df.prior = 0, ...)

Arguments

fit

object from lmFitResponses(), glmFitResponses(), lmmFitResponses(), glmmFitResponses()

coef

specify which coefficient, or combination of coefficients to test. Can be 1) a single coef name to test one coefficient: "x1", 2) an array coef names to test multiple coefficients: c("x1", "x2"), 3) string of linear functions of coefficients: "x1 - x2".

ddf

"finite": use Satterthwaite approximation to denominator degrees of freedom for the Student t or F distribution, or "asymptotic" to use normal or chisq distribution as null for the test statistic

...

other args, not used here

df.prior

augment degrees of freedom due to shrinking residual variance terms. Defaults to zero to have no impact

Value

tibble storing hypothesis test for each response

See also

car::linearHypothesis()

Examples

# simulate data
n <- 1000
m <- 3
nc <- 2
Y <- matrix(rnorm(n * m), m, n)
data = data.frame(x = sample(c("0", "1", "2"), n, replace=TRUE))
rownames(Y) <- seq(m) 
X <- model.matrix(~ x, data)

# fit regressions with model j using Y[,j] as a response
fit <- lmFitResponses(Y, X, detail=2)

# Univariate test
results(fit, "x1")
#> # A tibble: 3 × 5
#>   ID    Estimate     se    df p.value
#>   <chr>    <dbl>  <dbl> <dbl>   <dbl>
#> 1 1      -0.0183 0.0798   997  0.819 
#> 2 2      -0.0406 0.0790   997  0.608 
#> 3 3       0.139  0.0816   997  0.0881

# Joint test of 2 coefs
results(fit, c("x1", "x2"))
#> # A tibble: 3 × 5
#>   ID         F   df1   df2 p.value
#>   <chr>  <dbl> <int> <dbl>   <dbl>
#> 1 1     0.0271     2   997   0.973
#> 2 2     1.58       2   997   0.207
#> 3 3     1.47       2   997   0.230

# Contrast to compare coefficeints
results(fit, c("x1 - x2"))
#> # A tibble: 3 × 5
#>   ID    Estimate     se    df p.value
#>   <chr>    <dbl>  <dbl> <dbl>   <dbl>
#> 1 1      -0.0119 0.0809   997   0.883
#> 2 2       0.0992 0.0801   997   0.216
#> 3 3       0.0565 0.0828   997   0.495