Table printing and column width/adjustment

[strengejacke]

@DominiqueMakowski @mattansb @bwiernik @IndrajeetPatil

I have revised the function of the width argument in insight::export_table() (see examples), which now allows the alignment of (multiple) tables in model_parameters() as well.

library(parameters)
library(lme4)
#> Loading required package: Matrix

data("qol_cancer")
qol_cancer <- cbind(
  qol_cancer,
  datawizard::demean(qol_cancer, select = c("phq4", "QoL"), group = "ID")
)

model1 <- lmer(Sepal.Width ~ Petal.Length + (1|Species), data = iris)
model2 <- lmer(
  QoL ~ time + phq4_within + phq4_between + (1 | ID),
  data = qol_cancer
)


model_parameters(model1, effects = "all")
#> # Fixed Effects
#> 
#> Parameter    | Coefficient |   SE |       95% CI | t(146) |      p
#> ------------------------------------------------------------------
#> (Intercept)  |        2.00 | 0.56 | [0.89, 3.11] |   3.56 | < .001
#> Petal Length |        0.28 | 0.06 | [0.16, 0.40] |   4.75 | < .001
#> 
#> # Random Effects
#> 
#> Parameter               | Coefficient
#> -------------------------------------
#> SD (Intercept: Species) |        0.89
#> SD (Residual)           |        0.32
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

model_parameters(model1, effects = "all") |> print(column_width = "fixed")
#> # Fixed Effects
#> 
#> Parameter               | Coefficient |   SE |       95% CI | t(146) |      p
#> -----------------------------------------------------------------------------
#> (Intercept)             |        2.00 | 0.56 | [0.89, 3.11] |   3.56 | < .001
#> Petal Length            |        0.28 | 0.06 | [0.16, 0.40] |   4.75 | < .001
#> 
#> # Random Effects
#> 
#> Parameter               | Coefficient
#> -------------------------------------
#> SD (Intercept: Species) |        0.89
#> SD (Residual)           |        0.32
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

model_parameters(model2, effects = "fixed")
#> Parameter   | Coefficient |   SE |         95% CI | t(558) |      p
#> -------------------------------------------------------------------
#> (Intercept) |       71.53 | 1.56 | [68.48, 74.59] |  45.98 | < .001
#> time        |        1.09 | 0.64 | [-0.17,  2.34] |   1.70 | 0.089 
#> 
#> # Within-Effects
#> 
#> Parameter   | Coefficient |   SE |         95% CI | t(558) |      p
#> -------------------------------------------------------------------
#> phq4 within |       -3.66 | 0.41 | [-4.46, -2.86] |  -8.95 | < .001
#> 
#> # Between-Effects
#> 
#> Parameter    | Coefficient |   SE |         95% CI | t(558) |      p
#> --------------------------------------------------------------------
#> phq4 between |       -6.28 | 0.50 | [-7.27, -5.30] | -12.53 | < .001
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

model_parameters(model2, effects = "fixed") |> print(column_width = "fixed")
#> Parameter    | Coefficient |   SE |         95% CI | t(558) |      p
#> --------------------------------------------------------------------
#> (Intercept)  |       71.53 | 1.56 | [68.48, 74.59] |  45.98 | < .001
#> time         |        1.09 | 0.64 | [-0.17,  2.34] |   1.70 |  0.089
#> 
#> # Within-Effects
#> 
#> Parameter    | Coefficient |   SE |         95% CI | t(558) |      p
#> --------------------------------------------------------------------
#> phq4 within  |       -3.66 | 0.41 | [-4.46, -2.86] |  -8.95 | < .001
#> 
#> # Between-Effects
#> 
#> Parameter    | Coefficient |   SE |         95% CI | t(558) |      p
#> --------------------------------------------------------------------
#> phq4 between |       -6.28 | 0.50 | [-7.27, -5.30] | -12.53 | < .001
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

model_parameters(model2, effects = "fixed") |> 
  print(column_width = c(Parameter = 15, SE = 7, `95% CI` = 18))
#> Parameter       | Coefficient |      SE |             95% CI | t(558) |      p
#> ------------------------------------------------------------------------------
#> (Intercept)     |       71.53 |    1.56 |     [68.48, 74.59] |  45.98 | < .001
#> time            |        1.09 |    0.64 |     [-0.17,  2.34] |   1.70 | 0.089 
#> 
#> # Within-Effects
#> 
#> Parameter       | Coefficient |      SE |             95% CI | t(558) |      p
#> ------------------------------------------------------------------------------
#> phq4 within     |       -3.66 |    0.41 |     [-4.46, -2.86] |  -8.95 | < .001
#> 
#> # Between-Effects
#> 
#> Parameter       | Coefficient |      SE |             95% CI | t(558) |      p
#> ------------------------------------------------------------------------------
#> phq4 between    |       -6.28 |    0.50 |     [-7.27, -5.30] | -12.53 | < .001
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

[strengejacke]

sep and ci_brackets are now also passed down to insight::export_table() and insight::format_table():

library(parameters)
library(lme4)
#> Loading required package: Matrix

data("qol_cancer")
qol_cancer <- cbind(
  qol_cancer,
  datawizard::demean(qol_cancer, select = c("phq4", "QoL"), group = "ID")
)
model2 <- lmer(
  QoL ~ time + phq4_within + phq4_between + (1 | ID),
  data = qol_cancer
)

model_parameters(model2, effects = "fixed") |> 
  print(column_width = "fixed", 
        sep = "   ",
        ci_brackets = FALSE)
#> Parameter      Coefficient     SE         95% CI   t(558)        p
#> ------------------------------------------------------------------
#> (Intercept)          71.53   1.56   68.48, 74.59    45.98   < .001
#> time                  1.09   0.64   -0.17,  2.34     1.70    0.089
#> 
#> # Within-Effects
#> 
#> Parameter      Coefficient     SE         95% CI   t(558)        p
#> ------------------------------------------------------------------
#> phq4 within          -3.66   0.41   -4.46, -2.86    -8.95   < .001
#> 
#> # Between-Effects
#> 
#> Parameter      Coefficient     SE         95% CI   t(558)        p
#> ------------------------------------------------------------------
#> phq4 between         -6.28   0.50   -7.27, -5.30   -12.53   < .001
#> 
#> Uncertainty intervals (equal-tailed) and p values (two-tailed) computed using a
#>   Wald t-distribution approximation.

^{Created on 2021-11-11 by the reprex package (v2.0.1)}

[DominiqueMakowski]

very impressive!

[IndrajeetPatil]

So pretty! Thanks, Daniel 🧙

[strengejacke]

And it also renders nicely in rmarkdown presentations.
Here's a demo of three different ways to produce text tables and markdown and HTML in a presentation (from https://github.com/strengejacke/UKEtemplate):

[bwiernik]

Wow!