Include recommendation to use meta-package

easystats/easystats#308

R/estimate_grouplevel.R

@@ -47,7 +47,7 @@
 #' # Bayesian models
 #' \donttest{
 #' if (require("rstanarm")) {
-#'   model <- rstanarm::stan_lmer(mpg ~ hp + (1 | carb) + (1|gear), data = mtcars, refresh = 0)
+#'   model <- rstanarm::stan_lmer(mpg ~ hp + (1 | carb) + (1 | gear), data = mtcars, refresh = 0)
 #'   # Broken estimate_grouplevel(model)
 #' }
 #' }

README.Rmd

@@ -54,17 +54,20 @@ Development | R-universe | `install.packages("modelbased", repos = "https://easy
 
 Once you have downloaded the package, you can then load it using:
 
-  ```{r, eval=FALSE}
+```{r, eval=FALSE}
 library("modelbased")
 ```
 
-## Documentation
+> **Tip**
+>
+> **Instead of `library(datawizard)`, use `library(easystats)`.**
+> **This will make all features of the  easystats-ecosystem available.**
+>
+> **To stay updated, use `easystats::install_latest()`.**
 
-[![Documentation](https://img.shields.io/badge/documentation-modelbased-orange.svg?colorB=E91E63)](https://easystats.github.io/modelbased/)
-[![Blog](https://img.shields.io/badge/blog-easystats-orange.svg?colorB=FF9800)](https://easystats.github.io/blog/posts/)
-[![Features](https://img.shields.io/badge/features-modelbased-orange.svg?colorB=2196F3)](https://easystats.github.io/modelbased/reference/index.html)
+## Documentation
 
-Click on the buttons above to access the package [**documentation**](https://easystats.github.io/modelbased/) and the [**easystats blog**](https://easystats.github.io/blog/posts/), and check-out these vignettes:
+Access the package [**documentation**](https://easystats.github.io/modelbased/), and check-out these vignettes:
 
 - [**Visualisation matrix**](https://easystats.github.io/modelbased/articles/visualisation_matrix.html)
 - [**Marginal means**](https://easystats.github.io/modelbased/articles/estimate_means.html)

README.md

@@ -5,6 +5,7 @@
 [![downloads](http://cranlogs.r-pkg.org/badges/modelbased)](https://cran.r-project.org/package=modelbased)
 [![total](https://cranlogs.r-pkg.org/badges/grand-total/modelbased)](https://cranlogs.r-pkg.org/)
 [![status](https://tinyverse.netlify.com/badge/modelbased)](https://CRAN.R-project.org/package=modelbased)
+[![lifecycle](https://img.shields.io/badge/lifecycle-maturing-blue.svg)](https://lifecycle.r-lib.org/articles/stages.html)
 
 ***Taking your models to new heights***
 
@@ -36,64 +37,65 @@ Once you have downloaded the package, you can then load it using:
 library("modelbased")
 ```
 
-## Documentation
+> **Tip**
+>
+> **Instead of `library(datawizard)`, use `library(easystats)`.** **This
+> will make all features of the easystats-ecosystem available.**
+>
+> **To stay updated, use `easystats::install_latest()`.**
 
-[![Documentation](https://img.shields.io/badge/documentation-modelbased-orange.svg?colorB=E91E63)](https://easystats.github.io/modelbased/)
-[![Blog](https://img.shields.io/badge/blog-easystats-orange.svg?colorB=FF9800)](https://easystats.github.io/blog/posts/)
-[![Features](https://img.shields.io/badge/features-modelbased-orange.svg?colorB=2196F3)](https://easystats.github.io/modelbased/reference/index.html)
+## Documentation
 
-Click on the buttons above to access the package
-[**documentation**](https://easystats.github.io/modelbased/) and the
-[**easystats blog**](https://easystats.github.io/blog/posts/), and
+Access the package
+[**documentation**](https://easystats.github.io/modelbased/), and
 check-out these vignettes:
 
--   [**Visualisation
-    matrix**](https://easystats.github.io/modelbased/articles/visualisation_matrix.html)
--   [**Marginal
-    means**](https://easystats.github.io/modelbased/articles/estimate_means.html)
--   [**Contrast
-    analysis**](https://easystats.github.io/modelbased/articles/estimate_contrasts.html)
--   [**Marginal
-    effects**](https://easystats.github.io/modelbased/articles/estimate_slopes.html)
--   [**Use a model to make
-    predictions**](https://easystats.github.io/modelbased/articles/estimate_response.html)
--   [**Describe non-linear
-    curves**](https://easystats.github.io/modelbased/articles/describe_nonlinear.html)
--   [**Interpret models using Effect
-    Derivatives**](https://easystats.github.io/modelbased/articles/derivatives.html)
--   [**Estimate and re-use random
-    effects**](https://easystats.github.io/modelbased/articles/estimate_grouplevel.html)
--   [**The modelisation
-    approach**](https://easystats.github.io/modelbased/articles/modelisation_approach.html)
+- [**Visualisation
+  matrix**](https://easystats.github.io/modelbased/articles/visualisation_matrix.html)
+- [**Marginal
+  means**](https://easystats.github.io/modelbased/articles/estimate_means.html)
+- [**Contrast
+  analysis**](https://easystats.github.io/modelbased/articles/estimate_contrasts.html)
+- [**Marginal
+  effects**](https://easystats.github.io/modelbased/articles/estimate_slopes.html)
+- [**Use a model to make
+  predictions**](https://easystats.github.io/modelbased/articles/estimate_response.html)
+- [**Describe non-linear
+  curves**](https://easystats.github.io/modelbased/articles/describe_nonlinear.html)
+- [**Interpret models using Effect
+  Derivatives**](https://easystats.github.io/modelbased/articles/derivatives.html)
+- [**Estimate and re-use random
+  effects**](https://easystats.github.io/modelbased/articles/estimate_grouplevel.html)
+- [**The modelisation
+  approach**](https://easystats.github.io/modelbased/articles/modelisation_approach.html)
 
 # Features
 
 The package is built around 5 main functions:
 
--   [`estimate_means()`](https://easystats.github.io/modelbased/reference/estimate_means.html):
-    Estimates the average values at each factor levels
--   [`estimate_contrasts()`](https://easystats.github.io/modelbased/reference/estimate_contrasts.html):
-    Estimates and tests contrasts between different factor levels
--   [`estimate_slopes()`](https://easystats.github.io/modelbased/reference/estimate_slopes.html):
-    Estimates the slopes of numeric predictors at different factor
-    levels or alongside a numeric predictor
--   [`estimate_expectation()`](https://easystats.github.io/modelbased/articles/estimate_response.html):
-    Predict the response variable using the model
+- [`estimate_means()`](https://easystats.github.io/modelbased/reference/estimate_means.html):
+  Estimates the average values at each factor levels
+- [`estimate_contrasts()`](https://easystats.github.io/modelbased/reference/estimate_contrasts.html):
+  Estimates and tests contrasts between different factor levels
+- [`estimate_slopes()`](https://easystats.github.io/modelbased/reference/estimate_slopes.html):
+  Estimates the slopes of numeric predictors at different factor levels
+  or alongside a numeric predictor
+- [`estimate_expectation()`](https://easystats.github.io/modelbased/articles/estimate_response.html):
+  Predict the response variable using the model
 
 These functions are powered by the
 [`visualisation_matrix()`](https://easystats.github.io/modelbased/reference/visualisation_matrix.html)
 function, a smart tool for guessing the appropriate reference grid.
 
 ## Create smart grids to represent complex interactions
 
--   **Problem**: I want to graphically represent the interaction between
-    two continuous variable. On top of that, I would like to express one
-    of them in terms of standardized change (i.e., standard deviation
-    relative to the mean).
--   **Solution**: Create a data grid following the desired
-    specifications, and feed it to the model to obtain predictions.
-    Format some of the columns for better readability, and plot using
-    `ggplot`.
+- **Problem**: I want to graphically represent the interaction between
+  two continuous variable. On top of that, I would like to express one
+  of them in terms of standardized change (i.e., standard deviation
+  relative to the mean).
+- **Solution**: Create a data grid following the desired specifications,
+  and feed it to the model to obtain predictions. Format some of the
+  columns for better readability, and plot using `ggplot`.
 
 Check-out [**this
 vignette**](https://easystats.github.io/modelbased/articles/visualisation_matrix.html)
@@ -135,12 +137,12 @@ ggplot(iris, aes(x = Petal.Length, y = Sepal.Length)) +
 
 ## Estimate marginal means
 
--   **Problem**: My model has a factor as a predictor, and the
-    parameters only return the difference between levels and the
-    intercept. I want to see the values *at* each factor level.
--   **Solution**: Estimate model-based means (“marginal means”). You can
-    visualize them by plotting their confidence interval and the
-    original data.
+- **Problem**: My model has a factor as a predictor, and the parameters
+  only return the difference between levels and the intercept. I want to
+  see the values *at* each factor level.
+- **Solution**: Estimate model-based means (“marginal means”). You can
+  visualize them by plotting their confidence interval and the original
+  data.
 
 Check-out [**this
 vignette**](https://easystats.github.io/modelbased/articles/estimate_means.html)
@@ -186,12 +188,12 @@ ggplot(iris, aes(x = Species, y = Sepal.Width)) +
 
 ## Contrast analysis
 
--   **Problem**: The parameters of my model only return the difference
-    between some of the factor levels and the intercept. I want to see
-    the differences between each levels, as I would do with post-hoc
-    comparison tests in ANOVAs.
--   **Solution**: Estimate model-based contrasts (“marginal contrasts”).
-    You can visualize them by plotting their confidence interval.
+- **Problem**: The parameters of my model only return the difference
+  between some of the factor levels and the intercept. I want to see the
+  differences between each levels, as I would do with post-hoc
+  comparison tests in ANOVAs.
+- **Solution**: Estimate model-based contrasts (“marginal contrasts”).
+  You can visualize them by plotting their confidence interval.
 
 Check-out [**this
 vignette**](https://easystats.github.io/modelbased/articles/estimate_contrasts.html)
@@ -210,7 +212,7 @@ contrasts
 ## ------------------------------------------------------------------------------
 ## setosa     | versicolor |       0.66 | [ 0.49,  0.82] | 0.07 |   9.69 | < .001
 ## setosa     |  virginica |       0.45 | [ 0.29,  0.62] | 0.07 |   6.68 | < .001
-## versicolor |  virginica |      -0.20 | [-0.37, -0.04] | 0.07 |  -3.00 | 0.009 
+## versicolor |  virginica |      -0.20 | [-0.37, -0.04] | 0.07 |  -3.00 | 0.003 
 ## 
 ## Marginal contrasts estimated at Species
 ## p-value adjustment method: Holm (1979)
@@ -220,13 +222,13 @@ contrasts
 
 ## Check the contrasts at different points of another linear predictor
 
--   **Problem**: In the case of an interaction between a factor and a
-    continuous variable, you might be interested in computing how the
-    differences between the factor levels (the contrasts) change
-    depending on the other continuous variable.
--   **Solution**: You can estimate the marginal contrasts at different
-    values of a continuous variable (the *modulator*), and plot these
-    differences (they are significant if their 95% CI doesn’t cover 0).
+- **Problem**: In the case of an interaction between a factor and a
+  continuous variable, you might be interested in computing how the
+  differences between the factor levels (the contrasts) change depending
+  on the other continuous variable.
+- **Solution**: You can estimate the marginal contrasts at different
+  values of a continuous variable (the *modulator*), and plot these
+  differences (they are significant if their 95% CI doesn’t cover 0).
 
 ``` r
 model <- lm(Sepal.Width ~ Species * Petal.Length, data = iris)
@@ -238,13 +240,13 @@ estimate_contrasts(model, at = "Petal.Length", length = 3)
 ## --------------------------------------------------------------------------------------------
 ## setosa     | versicolor |         1.00 |       1.70 | [ 0.87, 2.53] | 0.34 |   4.97 | < .001
 ## setosa     | versicolor |         3.95 |       1.74 | [ 0.16, 3.32] | 0.65 |   2.67 | 0.023 
-## setosa     | versicolor |         6.90 |       1.78 | [-1.71, 5.26] | 1.44 |   1.24 | 0.434 
-## setosa     |  virginica |         1.00 |       1.34 | [ 0.38, 2.30] | 0.40 |   3.38 | 0.003 
-## setosa     |  virginica |         3.95 |       1.79 | [ 0.19, 3.40] | 0.66 |   2.70 | 0.021 
-## setosa     |  virginica |         6.90 |       2.25 | [-1.19, 5.69] | 1.42 |   1.58 | 0.257 
-## versicolor |  virginica |         1.00 |      -0.36 | [-1.55, 0.83] | 0.49 |  -0.73 | 0.747 
-## versicolor |  virginica |         3.95 |       0.06 | [-0.30, 0.42] | 0.15 |   0.37 | 0.926 
-## versicolor |  virginica |         6.90 |       0.47 | [-0.22, 1.16] | 0.28 |   1.65 | 0.229 
+## setosa     | versicolor |         6.90 |       1.78 | [-1.71, 5.26] | 1.44 |   1.24 | 0.304 
+## setosa     |  virginica |         1.00 |       1.34 | [ 0.38, 2.30] | 0.40 |   3.38 | 0.002 
+## setosa     |  virginica |         3.95 |       1.79 | [ 0.19, 3.40] | 0.66 |   2.70 | 0.023 
+## setosa     |  virginica |         6.90 |       2.25 | [-1.19, 5.69] | 1.42 |   1.58 | 0.304 
+## versicolor |  virginica |         1.00 |      -0.36 | [-1.55, 0.83] | 0.49 |  -0.73 | 0.468 
+## versicolor |  virginica |         3.95 |       0.06 | [-0.30, 0.42] | 0.15 |   0.37 | 0.710 
+## versicolor |  virginica |         6.90 |       0.47 | [-0.22, 1.16] | 0.28 |   1.65 | 0.304 
 ## 
 ## Marginal contrasts estimated at Species
 ## p-value adjustment method: Holm (1979)
@@ -272,14 +274,14 @@ ggplot(contrasts, aes(x = Petal.Length, y = Difference, )) +
 
 ## Generate predictions from your model to compare it with original data
 
--   **Problem**: You fitted different models, and you want to
-    intuitively visualize how they compare in terms of fit quality and
-    prediction accuracy, so that you don’t only rely on abstract indices
-    of performance.
--   **Solution**: You can predict the response variable from different
-    models and plot them against the original true response. The closest
-    the points are on the identity line (the diagonal), the closest they
-    are from a perfect fit.
+- **Problem**: You fitted different models, and you want to intuitively
+  visualize how they compare in terms of fit quality and prediction
+  accuracy, so that you don’t only rely on abstract indices of
+  performance.
+- **Solution**: You can predict the response variable from different
+  models and plot them against the original true response. The closest
+  the points are on the identity line (the diagonal), the closest they
+  are from a perfect fit.
 
 Check-out [**this
 vignette**](https://easystats.github.io/modelbased/articles/estimate_response.html)
@@ -328,10 +330,10 @@ ggplot(data = pred1, aes(x = Petal.Length, y = Predicted)) +
 
 ## Extract and format group-level random effects
 
--   **Problem**: You have a mixed model and you would like to easily
-    access the random part, i.e., the group-level effects (e.g., the
-    individuals scores).
--   **Solution**: You can apply `estimate_grouplevel` on a mixed model.
+- **Problem**: You have a mixed model and you would like to easily
+  access the random part, i.e., the group-level effects (e.g., the
+  individuals scores).
+- **Solution**: You can apply `estimate_grouplevel` on a mixed model.
 
 See [**this
 vignette**](https://easystats.github.io/modelbased/articles/estimate_grouplevel.html)
@@ -362,11 +364,11 @@ plot(random)
 
 ## Estimate derivative of non-linear relationships (e.g., in GAMs)
 
--   **Problem**: You model a non-linear relationship using polynomials,
-    splines or GAMs. You want to know which parts of the curve are
-    significant positive or negative trends.
--   **Solution**: You can estimate the *derivative* of smooth using
-    `estimate_slopes`.
+- **Problem**: You model a non-linear relationship using polynomials,
+  splines or GAMs. You want to know which parts of the curve are
+  significant positive or negative trends.
+- **Solution**: You can estimate the *derivative* of smooth using
+  `estimate_slopes`.
 
 The two plots below represent the modeled (non-linear) effect estimated
 by the model, i.e., the relationship between the outcome and the
@@ -403,12 +405,12 @@ see::plots(
 
 ## Describe the smooth term by its linear parts
 
--   **Problem**: You model a non-linear relationship using polynomials,
-    splines or GAMs. You want to describe it in terms of linear parts:
-    where does it decrease, how much, where does it increase, etc.
--   **Solution**: You can apply `describe_nonlinear` on a predicted
-    relationship that will return the different parts of increase and
-    decrease.
+- **Problem**: You model a non-linear relationship using polynomials,
+  splines or GAMs. You want to describe it in terms of linear parts:
+  where does it decrease, how much, where does it increase, etc.
+- **Solution**: You can apply `describe_nonlinear` on a predicted
+  relationship that will return the different parts of increase and
+  decrease.
 
 See [**this
 vignette**](https://easystats.github.io/modelbased/articles/describe_nonlinear.html)

tests/testthat/test-estimate_means.R

@@ -1,8 +1,8 @@
 if (require("testthat") &&
-    require("modelbased") &&
-    require("rstanarm") &&
-    require("insight") &&
-    require("emmeans")) {
+  require("modelbased") &&
+  require("rstanarm") &&
+  require("insight") &&
+  require("emmeans")) {
   test_that("estimate_means", {
     data <- mtcars
     data$gear <- as.factor(data$gear)

tests/testthat/test-estimate_slopes.R

@@ -1,8 +1,8 @@
 if (require("testthat") &&
-    require("modelbased") &&
-    require("rstanarm") &&
-    require("insight") &&
-    require("emmeans")) {
+  require("modelbased") &&
+  require("rstanarm") &&
+  require("insight") &&
+  require("emmeans")) {
   test_that("estimate_slopes", {
     model <- lm(Sepal.Width ~ Species * Petal.Length, data = iris)