toffee_tbl: Create a nicely formatted table from a model object.
In seankross/toffee: Create Tables for Easy Understanding of Model Objects

Description Usage Arguments Value Examples

Given a model object (like from stats::lm(), stats::glm(), or lme4::glmer()) this function makes a table intended for interactive use or publication. The model coefficients and confidence intervals displayed in the table have had their inverse link function applied, therefore they are in the units of the response variable.

1 2	toffee_tbl(model, conf_level = 0.95, digits = 2, concat_signif = TRUE, odds_ratio = TRUE, ci_fmt = TRUE, ...)

`model`	A model object that is compatible with `broom::tidy()`.
`conf_level`	The confidence level that will be used for computing the confidence interval (for example: `0.95` or `0.99`).
`digits`	The number of decimal places that numbers in the table should be rounded to. For no rounding use `Inf`. This argument is provided to `base::round()`. The default value is `2`.
`concat_signif`	If `TRUE` the resulting data frame will append significance symbols as generated by `toffee_signif()`. If `FALSE` the significance symbols will appear in a separate column called `Significant`.
`odds_ratio`	If `TRUE` and the family of the model provided is `logit` then the model coefficients and confidence intervals returned in the resulting table will be odds ratios.
`ci_fmt`	If `TRUE` then the confidence interval will be formatted between brackets in one column called `CI`. If `FALSE` the lower and upper bounds of the confidence interval will be in their own columns called `Lower_CI` and `Upper_CI` respectfully.
`...`	Arguments that will be passed to `toffee_signif()`.

A tibble::tibble() with the following columns: the name of the coefficient in the model (Variable), the value of the coefficient with inverse of the link function applied (Estimate or Odds_Ratio), the confidence interval for the coefficient (CI), the p-value for the estimate, (p_value), and optionally symbols representing levels of significance (Significant).

library(dplyr)
library(toffee)

linear_model <- toffee_forest %>%
  lm(Volume ~ Girth + Height + Branches, data = .)

# Basic usage
linear_model %>%
  toffee_tbl()

 # # A tibble: 4 x 4
 #   Variable    Estimate CI               p_value
 #   <chr>          <dbl> <chr>            <chr>
 # 1 (Intercept)   -67.2  [-86.81, -47.59] < 0.01***
 # 2 Girth           4.75 [4.23, 5.27]     < 0.01***
 # 3 Height          0.33 [0.08, 0.59]     0.01*
 # 4 Branches        0.46 [-0.03, 0.95]    0.07

# Using the `chars` argument from toffee_signif
linear_model %>%
  toffee_tbl(thresholds = 0.01, chars = "*")

 # # A tibble: 4 x 4
 #   Variable    Estimate CI               p_value
 #   <chr>          <dbl> <chr>            <chr>
 # 1 (Intercept)   -67.2  [-86.81, -47.59] < 0.01*
 # 2 Girth           4.75 [4.23, 5.27]     < 0.01*
 # 3 Height          0.33 [0.08, 0.59]     0.01
 # 4 Branches        0.46 [-0.03, 0.95]    0.07

# Separating the significance symbols into their own column
linear_model %>%
  toffee_tbl(concat_signif = FALSE)

 # # A tibble: 4 x 5
 #   Variable    Estimate CI               p_value Significant
 #   <chr>          <dbl> <chr>            <chr>   <chr>
 # 1 (Intercept)   -67.2  [-86.81, -47.59] < 0.01  ***
 # 2 Girth           4.75 [4.23, 5.27]     < 0.01  ***
 # 3 Height          0.33 [0.08, 0.59]     0.01    *
 # 4 Branches        0.46 [-0.03, 0.95]    0.07    ""

logistic_model <- toffee_forest %>%
  glm(Healthy ~ Girth + Height + Branches + Volume, data = .,
      family = binomial())

logistic_model %>%
  toffee_tbl()

 # # A tibble: 5 x 4
 #   Variable    Odds_Ratio CI                  p_value
 #   <chr>            <dbl> <chr>               <chr>
 # 1 (Intercept)       0.05 [0, 10626998633.06] 0.81
 # 2 Girth             0.75 [0.14, 3.65]        0.72
 # 3 Height            0.92 [0.73, 1.14]        0.45
 # 4 Branches          2    [1.31, 3.82]        0.01**
 # 5 Volume            1.03 [0.75, 1.45]        0.86