### Install interactions Package

Source: https://github.com/jacob-long/interactions/blob/master/README.md

Install the interactions package from CRAN. This is the first step before using any of its functions.

```r
install.packages("interactions")
```

--------------------------------

### Install and Load Interactions Package

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/00-START-HERE.md

Installs the interactions package from CRAN and loads it into the R session. This is the first step before using any functions from the package.

```r
install.packages("interactions")
library(interactions)
```

--------------------------------

### Install Interactions Package Development Version

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/README.md

Install the latest development version of the interactions package directly from GitHub using the devtools package.

```r
devtools::install_github("jacob-long/interactions")
```

--------------------------------

### Get Help for Functions and Vignettes in R

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Use the '?' operator to access help for specific functions like sim_slopes and interact_plot. To explore detailed examples and workflows, use vignette() and browseVignettes() for the 'interactions' package.

```r
# Help for a function
?sim_slopes
?interact_plot

# Vignettes
vignette("interactions")

# List vignettes
browseVignettes("interactions")
```

--------------------------------

### Probe Interaction with Survey Data

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This example demonstrates using probe_interaction with survey-weighted data using the 'survey' package. It shows how to set up the survey design and fit a survey-weighted generalized linear model.

```r
library(survey)
data(api)
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)
regmodel <- svyglm(api00 ~ ell * meals, design = dstrat)

results <- probe_interaction(regmodel, pred = ell, modx = meals)
results
```

--------------------------------

### Package Documentation File Structure

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

This tree outlines the organization of documentation files within the package, indicating the purpose of each file from the README to specific API references and usage guides.

```bash
output/
├── README.md                          # Start here
├── INDEX.md                           # This file
├── api-reference-sim_slopes.md        # Function docs
├── api-reference-interact_plot.md
├── api-reference-johnson_neyman.md
├── api-reference-sim_margins.md
├── api-reference-cat_plot.md
├── api-reference-probe_interaction.md
├── parameter-reference.md             # Parameter guide
├── s3-methods.md                      # S3 method docs
├── types.md                           # Object structures
└── usage-guide.md                     # Examples & workflows
```

--------------------------------

### Linear Model Example

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

Example of fitting a linear model with an interaction term.

```r
model <- lm(y ~ x * z, data = data)
```

--------------------------------

### Install and Load Interactions Package

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Installs the interactions package from CRAN and loads it along with ggplot2 and dplyr for data manipulation and plot customization.

```r
# Install from CRAN
install.packages("interactions")

# Load package
library(interactions)
library(ggplot2)  # For plot customization
library(dplyr)    # For data manipulation
```

--------------------------------

### Probe Interaction with Custom Moderator Values

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This example demonstrates how to specify exact values for the moderator variable (modx) when using probe_interaction. This allows for a more targeted analysis of the interaction effect at specific points.

```r
# Specify exact moderator values
results <- probe_interaction(fit, pred = hp, modx = wt,
                             modx.values = c(2, 3, 4, 5))
results
```

--------------------------------

### Survey Model Example

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

Example of fitting a survey-weighted generalized linear model with an interaction term.

```r
model <- svyglm(y ~ x * z, design = survey_design)
```

--------------------------------

### Probe Interaction with Factor Moderator

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This example demonstrates using probe_interaction when the moderator variable is a factor (categorical). It also includes plotting the points for clarity.

```r
fit_cat <- lm(Petal.Length ~ Petal.Width * Species, data = iris)
results <- probe_interaction(fit_cat, pred = Petal.Width, modx = Species,
                             plot.points = TRUE)
results
```

--------------------------------

### Simple Slopes for Survey Design Models

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_slopes.md

Analyze simple slopes for models fitted with survey data using the 'survey' package. This example demonstrates fitting a 'svyglm' model and then applying sim_slopes to analyze the interaction between 'ell' and 'meals'.

```r
library(survey)
data(api)
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)
regmodel <- svyglm(api00 ~ ell * meals, design = dstrat)

results <- sim_slopes(regmodel, pred = ell, modx = meals)
results
```

--------------------------------

### Probe Interaction with Visualization Options

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This example shows how to enhance the probe_interaction plot by adding points, confidence intervals, and visualizing the Johnson-Neyman interval. It requires setting specific arguments like plot.points, interval, and jnplot.

```r
# Add points and intervals to plot, plot J-N interval
results <- probe_interaction(fit, pred = hp, modx = wt, 
                             plot.points = TRUE, 
                             interval = TRUE,
                             jnplot = TRUE)
results
```

--------------------------------

### Three-Way Interaction Analysis

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_slopes.md

Examine simple slopes in a three-way interaction model. This example fits a model predicting 'Income' from 'Frost', 'Murder', and 'Illiteracy' and then uses sim_slopes to analyze the effect of 'Murder' at specific values of 'Illiteracy' and 'Frost'.

```r
states <- as.data.frame(state.x77)
fit3 <- lm(Income ~ Frost * Murder * Illiteracy, data = states)

# Analyze slopes at specific values
results <- sim_slopes(fit3, pred = Murder, modx = Illiteracy, mod2 = Frost,
                      modx.values = c(5, 10, 15), mod2.values = c(0, 100))
results
```

--------------------------------

### Logistic Model Example

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

Example of fitting a logistic regression model with an interaction term.

```r
model <- glm(y ~ x * z, data = data, family = binomial)
```

--------------------------------

### Tidying Sim Slopes Results with dplyr

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

This example demonstrates how to convert the results of `sim_slopes` into a tidy data frame using the `tidy` S3 method and then filter and arrange the results using `dplyr`.

```r
slopes <- sim_slopes(model, pred = x, modx = z)

# Convert to tidy format
tidy_slopes <- tidy(slopes)

# Use with dplyr
tidy_slopes %>%
  filter(p.value < 0.05) %>%
  arrange(estimate)
```

--------------------------------

### Basic Interaction Analysis Workflow

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

A template for fitting a linear model with an interaction term, analyzing the interaction using sim_slopes, and visualizing it with interact_plot or probe_interaction.

```r
# 1. Fit a model with interaction
model <- lm(outcome ~ predictor * moderator, data = your_data)

# 2. Analyze the interaction
analysis <- sim_slopes(model, pred = predictor, modx = moderator)
print(analysis)

# 3. Visualize the interaction
plot <- interact_plot(model, pred = predictor, modx = moderator)
print(plot)

# Or combine both:
combined <- probe_interaction(model, pred = predictor, modx = moderator)
print(combined)
```

--------------------------------

### Export sim_slopes results as a Word document

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Demonstrates converting `sim_slopes` results to a `huxtable` and exporting it as a .docx file. Requires `interactions` and `huxtable` packages. Specifies the number of digits for the output.

```r
library(interactions)
library(huxtable)

fit <- lm(mpg ~ hp * wt, data = mtcars)
slopes <- sim_slopes(fit, pred = hp, modx = wt, digits = 3)

# Export as formatted table
table <- as_huxtable(slopes)
export_table(table, file = "slopes_table.docx")
```

--------------------------------

### Basic Two-Way Interaction Analysis and Visualization

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Demonstrates fitting a linear model and then analyzing and visualizing a basic two-way interaction using sim_slopes and interact_plot.

```R
library(interactions)

# Fit model
model <- lm(mpg ~ hp * wt, data = mtcars)

# Analyze
sim_slopes(model, pred = hp, modx = wt)

# Visualize
interact_plot(model, pred = hp, modx = wt)

# Or both
probe_interaction(model, pred = hp, modx = wt)
```

--------------------------------

### Extract Number of Observations from sim_slopes

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Use `nobs()` to extract the total number of observations used in the simple slopes analysis. This is a direct way to get the sample size for the model.

```r
results <- sim_slopes(model, pred = x, modx = z)
nobs(results)
```

--------------------------------

### Probe interaction and visualize simultaneously

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/README.md

Use the probe_interaction function to both analyze and visualize interaction effects in a single step. This is a convenient shortcut for common analysis tasks.

```r
probe_interaction(fit, pred = x, modx = z)
```

--------------------------------

### Standardized Predictors for Interpretation

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_slopes.md

Use standardized predictors in the sim_slopes analysis for easier interpretation of regression coefficients. This example centers and standardizes all variables before analyzing simple slopes, setting the 'digits' argument to control output precision.

```r
# Center and standardize for regression interpretation
results <- sim_slopes(fiti, pred = hp, modx = wt,
                      centered = "all", digits = 3)
results
```

--------------------------------

### Accessing probe_interaction Components

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/types.md

Demonstrates how to access the simslopes table, the full simslopes object, and the interactplot object from a probe_interaction result. Also shows basic plot customization.

```r
results <- probe_interaction(model, pred = x, modx = z)

# Slopes table
slopes_table <- results$simslopes$slopes

# Full slopes object
slopes_obj <- results$simslopes

# Plot object
plot_obj <- results$interactplot

# Customize plot
results$interactplot + theme_minimal()
```

--------------------------------

### Plot Simple Margins

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Provides a basic plot method for sim_margins objects. The implementation is minimal.

```r
results <- sim_margins(model, pred = x, modx = z)
plot(results)
```

--------------------------------

### Three-Way Interaction Analysis and Visualization

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Shows how to fit a linear model for a three-way interaction and analyze/visualize it, including specifying moderator values and faceting the plot.

```R
# Model
states <- as.data.frame(state.x77)
model <- lm(Income ~ Murder * Illiteracy * Frost, data = states)

# Analyze at mod2 values
slopes <- sim_slopes(model, pred = Murder, modx = Illiteracy, 
                     mod2 = Frost, mod2.values = "plus-minus")

# Visualize with facetting
plot <- interact_plot(model, pred = Murder, modx = Illiteracy,
                      mod2 = Frost)
```

--------------------------------

### Basic Probe Interaction Usage

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This snippet demonstrates the basic usage of probe_interaction to analyze and plot interaction effects from a linear model. It shows how to fit a model, call probe_interaction, and access the separate analysis and plot components.

```r
library(interactions)

# Fit interaction model
fit <- lm(mpg ~ hp * wt, data = mtcars)

# Analyze and plot in one call
results <- probe_interaction(fit, pred = hp, modx = wt)

# Print slopes and display plot
results

# Extract components separately
slopes <- results$simslopes
plot <- results$interactplot
```

--------------------------------

### Combine tidy slopes and margins results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Illustrates combining tidy output from both `sim_slopes` and `sim_margins` using `dplyr::bind_rows`. Requires `interactions`, `dplyr`, and `bind_rows`.

```r
library(interactions)
library(dplyr)
library(bind_rows)

fit <- lm(mpg ~ hp * wt, data = mtcars)

# Compare slopes vs. margins
slopes_tidy <- tidy(sim_slopes(fit, pred = hp, modx = wt)) %>%
  mutate(method = "slopes")
margins_tidy <- tidy(sim_margins(fit, pred = hp, modx = wt)) %>%
  mutate(method = "margins")

combined <- bind_rows(slopes_tidy, margins_tidy)
```

--------------------------------

### Use Prediction Intervals

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-interact_plot.md

Illustrates how to display prediction intervals instead of confidence intervals by specifying int.type = "prediction" along with interval = TRUE.

```r
# Use prediction intervals instead
interact_plot(fit, pred = hp, modx = wt, interval = TRUE, int.type = "prediction")
```

--------------------------------

### Sensitivity Analysis: Different Centering Options

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Examine how different centering strategies ('all' vs. 'none') for predictors affect the estimated conditional intercepts. This helps in understanding the impact of centering on model interpretation.

```r
# All variables centered (default)
slopes_centered <- sim_slopes(model, pred = x, modx = z, 
                              centered = "all")

# No centering
slopes_none <- sim_slopes(model, pred = x, modx = z,
                          centered = "none")

# Compare how centering affects conditional intercepts
intercepts_comparison <- data.frame(
  centered = slopes_centered$ints$Est.,
  none = slopes_none$ints$Est.
)
print(intercepts_comparison)
```

--------------------------------

### Tidy sim_slopes results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Demonstrates tidying the results of `sim_slopes` into a data frame format for further analysis. Requires the `interactions` package.

```r
library(interactions)

fit <- lm(mpg ~ hp * wt, data = mtcars)
slopes <- sim_slopes(fit, pred = hp, modx = wt)

# Convert to tidy format for further analysis
tidy_slopes <- tidy(slopes)
head(tidy_slopes)
```

--------------------------------

### Simulate Slopes with Original Data

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/README.md

When using variable transformations in your formula (e.g., log(x)), provide the original data frame to the `data` argument in `sim_slopes` to ensure correct calculations.

```r
sim_slopes(fit, pred = y, modx = x, data = original_data)
```

--------------------------------

### Basic Margins Analysis with lm

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_margins.md

Computes marginal effects for a linear model with an interaction term at default moderator values. Requires the 'interactions' library.

```r
library(interactions)

# Fit interaction model
fit <- lm(mpg ~ hp * wt, data = mtcars)

# Compute margins at default moderator values
results <- sim_margins(fit, pred = hp, modx = wt)
results
```

--------------------------------

### Print Interaction Probe Results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Use this method to display both the simple slopes analysis table and the corresponding interaction plot. It combines tabular and visual summaries of interaction effects.

```r
results <- probe_interaction(model, pred = x, modx = z)
print(results)
```

--------------------------------

### Convert sim_margins to huxtable

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Converts simple margins results to a formatted huxtable object. Requires the `huxtable` package. The output is a huxtable object with a formatted table.

```r
results <- sim_margins(model, pred = x, modx = z)
table <- as_huxtable(results)
```

--------------------------------

### Simulate Marginal Effects Analysis

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Utilize sim_margins as an alternative to sim_slopes, employing a margins-based approach. It supports simulation and bootstrap standard errors and is beneficial for complex models.

```r
margins <- sim_margins(model, pred = x, modx = z)
```

--------------------------------

### Print Simple Margins Analysis Results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Use this method to display simple margins analysis results in a tabular format. It includes conditional effects, standard errors, confidence intervals, and p-values.

```r
results <- sim_margins(model, pred = x, modx = z)
print(results)
```

--------------------------------

### Basic Johnson-Neyman Interval with Plot

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-johnson_neyman.md

Computes the Johnson-Neyman interval and displays a plot of the results. Requires the 'interactions' package and a fitted linear model.

```r
library(interactions)

# Fit interaction model
states <- as.data.frame(state.x77)
states$HSGrad <- states$`HS Grad`
fit <- lm(Income ~ HSGrad + Murder * Illiteracy, data = states)

# Compute Johnson-Neyman interval
jn <- johnson_neyman(fit, pred = Murder, modx = Illiteracy, plot = TRUE)
jn
```

--------------------------------

### Compare Different Models with sim_slopes

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Compare the results of different regression models by applying sim_slopes to each and binding the results. This is useful for understanding how model complexity or additional predictors affect estimated slopes.

```r
# Model 1: Simple interaction
model1 <- lm(y ~ x * z, data = data)
slopes1 <- tidy(sim_slopes(model1, pred = x, modx = z)) %>%
  mutate(model = "Simple Interaction")

# Model 2: With additional predictor
model2 <- lm(y ~ x * z + w, data = data)
slopes2 <- tidy(sim_slopes(model2, pred = x, modx = z)) %>%
  mutate(model = "With Covariate")

# Compare
comparison <- bind_rows(slopes1, slopes2)
print(comparison)
```

--------------------------------

### Print Simple Slopes Analysis Results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Use this method to display formatted results from a simple slopes analysis. It shows conditional slopes, intercepts, and Johnson-Neyman intervals if requested.

```r
results <- sim_slopes(model, pred = x, modx = z)
print(results)  # or just results
```

--------------------------------

### Provide Original Data for Transformations

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

When your model formula includes transformations like poly(), sqrt(), or log(), you must provide the original data frame using the 'data' argument. This ensures correct calculations for transformed variables.

```r
model <- lm(y ~ poly(x, 2) * z, data = mydata)

interact_plot(model, pred = x, modx = z, data = mydata)
```

--------------------------------

### Bar Chart with Points for Interaction

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-cat_plot.md

Generates a bar chart for an interaction, with points overlaid and uncertainty intervals shown. This is an alternative visualization despite common criticisms of bar charts for interactions.

```r
# Some prefer this despite criticisms
p <- cat_plot(fit, pred = color, modx = cut, 
              geom = "bar", plot.points = TRUE, 
              interval = TRUE)
print(p)
```

--------------------------------

### Probe Interaction for Three-Way Interaction

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

This snippet illustrates how to use probe_interaction to analyze a three-way interaction effect. It requires specifying an additional moderator (mod2) and optionally its values.

```r
# Three-way interaction
states <- as.data.frame(state.x77)
fit3 <- lm(Income ~ Frost * Murder * Illiteracy, data = states)

results <- probe_interaction(fit3, pred = Murder, modx = Illiteracy,
                             mod2 = Frost, mod2.values = "plus-minus")
results
```

--------------------------------

### Extracting Results from probe_interaction

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-probe_interaction.md

Demonstrates how to access the individual components of the results object returned by probe_interaction for further analysis.

```r
results <- probe_interaction(fit, pred = x, modx = z)
slopes_table <- results$simslopes$slopes
plot_object <- results$interactplot
```

--------------------------------

### Simulate Slopes with Confidence Intervals and Robust Standard Errors

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

Use this snippet to simulate simple slopes from a model, displaying confidence intervals instead of standard errors and calculating robust standard errors (HC3).

```r
sim_slopes(model, pred = x, modx = z,
           confint = TRUE,     # Show CIs not SEs
           ci.width = 0.95,    # 95% CI
           robust = "HC3")     # Robust SEs
```

--------------------------------

### Survey Data Analysis with Interactions

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Perform interaction analysis on survey data using `svyglm` and the `survey` package. Demonstrates analysis with survey designs.

```r
library(survey)

data(api)

# Create survey design
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)

# Fit svyglm
model <- svyglm(api00 ~ ell * meals, design = dstrat)

# Analysis works the same way
slopes <- sim_slopes(model, pred = ell, modx = meals)
print(slopes)

plot <- interact_plot(model, pred = ell, modx = meals)
print(plot)
```

--------------------------------

### Survey Data Interaction Analysis

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Shows how to analyze an interaction within survey data using svyglm and sim_slopes, accounting for survey design complexities.

```R
library(survey)
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)
model <- svyglm(api00 ~ ell * meals, design = dstrat)

sim_slopes(model, pred = ell, modx = meals)
```

--------------------------------

### Simple Two-Way Interaction with Continuous Variables

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_slopes.md

Analyze simple slopes for a two-way interaction between continuous predictors. This snippet demonstrates fitting a linear model and then using sim_slopes to examine the effect of 'hp' at different values of 'wt'. The default analysis uses values of 'wt' at mean ± 1 standard deviation.

```r
library(interactions)

# Fit a model with interaction
fiti <- lm(mpg ~ hp * wt, data = mtcars)

# Analyze simple slopes at default values (mean ± 1 SD of wt)
results <- sim_slopes(fiti, pred = hp, modx = wt)
results
```

--------------------------------

### Analyze interaction effects with sim_slopes

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/README.md

Analyze the fitted model to probe the simple slopes of an interaction. Specify the focal predictor and the moderator variable.

```r
sim_slopes(fit, pred = x, modx = z)
```

--------------------------------

### Basic Interaction Plot

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-interact_plot.md

Generates a basic interaction plot for a linear model. Requires the interactions package and a fitted model object.

```r
library(interactions)

# Fit model
fit <- lm(mpg ~ hp * wt, data = mtcars)

# Plot interaction
p <- interact_plot(fit, pred = hp, modx = wt)
print(p)
```

--------------------------------

### Margins Analysis for Survey-Weighted Models

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_margins.md

Computes marginal effects for survey-weighted generalized linear models using the 'survey' package. Requires setting up a survey design object.

```r
library(survey)
data(api)
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)
regmodel <- svyglm(api00 ~ ell * meals, design = dstrat)

results <- sim_margins(regmodel, pred = ell, modx = meals)
results
```

--------------------------------

### Convert sim_slopes to huxtable

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Converts simple slopes results to a formatted huxtable object. Requires the `huxtable` package. The output is suitable for console printing or exporting to various file formats.

```r
results <- sim_slopes(model, pred = x, modx = z)
table <- as_huxtable(results)
print(table)
```

--------------------------------

### sim_slopes Function Signature

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_slopes.md

This is the complete function signature for sim_slopes, detailing all available parameters and their default values. It is useful for understanding the full range of options when performing simple slopes analysis.

```r
sim_slopes(
  model,
  pred,
  modx,
  mod2 = NULL,
  modx.values = NULL,
  mod2.values = NULL,
  centered = "all",
  at = NULL,
  data = NULL,
  cond.int = FALSE,
  johnson_neyman = TRUE,
  jnplot = FALSE,
  jnalpha = 0.05,
  robust = FALSE,
  digits = getOption("jtools-digits", default = 2),
  pvals = TRUE,
  confint = FALSE,
  ci.width = 0.95,
  cluster = NULL,
  modx.labels = NULL,
  mod2.labels = NULL,
  v.cov = NULL,
  v.cov.args = NULL,
  ...
)
```

--------------------------------

### Interaction Plot for Survey Data

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-interact_plot.md

Demonstrates plotting interactions for survey data using the survey package. Requires a svydesign object and a svyglm model.

```r
library(survey)
data(api)
dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw,
                    data = apistrat, fpc = ~fpc)
regmodel <- svyglm(api00 ~ ell * meals, design = dstrat)

p <- interact_plot(regmodel, pred = ell, modx = meals)
print(p)
```

--------------------------------

### Export Table to LaTeX

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Exports a summarized slopes table to a LaTeX file (.tex). Requires the huxtable package.

```r
# Export to LaTeX
export_table(table, file = "slopes_table.tex")
```

--------------------------------

### Accessing Object Components

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Extract specific components like slopes tables, Johnson-Neyman intervals, or fitted models from simulation results objects.

```r
results <- sim_slopes(fit, pred = x, modx = z)

# Access slopes table
slopes_table <- results$slopes

# Access Johnson-Neyman interval
jn_list <- results$jn

# Access all fitted models
models <- results$mods
```

--------------------------------

### Optimizing Computation Speed for Large Models

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Improve the performance of `sim_slopes` for large models by using simpler centering options or disabling the Johnson-Neyman interval calculation. These adjustments can significantly reduce computation time.

```r
# Use simpler centering to speed up
slopes <- sim_slopes(model, pred = x, modx = z,
                     centered = "none")  # Faster

# Or skip J-N interval
slopes <- sim_slopes(model, pred = x, modx = z,
                     johnson_neyman = FALSE)
```

--------------------------------

### Export Table to Excel

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Exports a summarized slopes table to an Excel file (.xlsx). Requires the huxtable package.

```r
# Export to Excel
export_table(table, file = "slopes_table.xlsx")
```

--------------------------------

### Logistic Regression Interaction Analysis and Visualization

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Demonstrates analyzing and visualizing interactions in a logistic regression model, showing slopes in log-odds and plotting on the probability scale.

```R
model <- glm(vs ~ hp * wt, data = mtcars, family = binomial)

# Slopes in log-odds
sim_slopes(model, pred = hp, modx = wt)

# Plot on probability scale
interact_plot(model, pred = hp, modx = wt, outcome.scale = "response")
```

--------------------------------

### Categorical Interaction Plot

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

Illustrates how to use cat_plot to visualize an interaction between two categorical variables, including confidence intervals.

```R
# Categorical × Categorical
model <- lm(price ~ cut * color, data = diamonds)
cat_plot(model, pred = cut, modx = color, interval = TRUE)
```

--------------------------------

### Probe Interaction with Plotting Options and Precision

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

This snippet probes interactions, showing data points, uncertainty intervals, and the J-N interval with a specified number of digits for precision.

```r
probe_interaction(model, pred = x, modx = z,
                  plot.points = TRUE,    # Show data
                  interval = TRUE,       # Show uncertainty
                  jnplot = TRUE,         # Show J-N interval
                  digits = 3)            # More precision
```

--------------------------------

### Tidy Simple Slopes Results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Converts the results of a sim_slopes analysis into a tidy data frame, compatible with the broom package. The output includes conditional slope estimates, standard errors, and p-values.

```r
results <- sim_slopes(model, pred = x, modx = z)
tidy_results <- tidy(results)
```

--------------------------------

### Overlay Observed Data with Jitter

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-interact_plot.md

Enhances the visibility of observed data points by adding jitter, which helps to reduce overplotting when many points share similar values.

```r
# With jitter to reduce overplotting
interact_plot(fit, pred = hp, modx = wt, plot.points = TRUE, jitter = 0.1)
```

--------------------------------

### Fit a model with an interaction

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/README.md

Fit a linear model including an interaction term between predictors x and z. This is a common first step for interaction analysis.

```r
fit <- lm(y ~ x * z, data = d)
```

--------------------------------

### Customize ggplot Object

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-cat_plot.md

Demonstrates how to customize the ggplot object returned by cat_plot using standard ggplot2 functions like theme_minimal and scale_color_manual.

```r
# Example customization
p <- cat_plot(model, pred = x, modx = z)
p + theme_minimal() + scale_color_manual(values = c("#E69F00", "#56B4E9"))
```

--------------------------------

### Export Table to Word

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/usage-guide.md

Exports a summarized slopes table to a Word document (.docx). Requires the huxtable package.

```r
library(huxtable)

slopes <- sim_slopes(model, pred = x, modx = z, digits = 3)

# Convert to huxtable
table <- as_huxtable(slopes)

# Export to Word
export_table(table, file = "slopes_table.docx")
```

--------------------------------

### Display Confidence Intervals Instead of Standard Errors

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/parameter-reference.md

Configure the output to display confidence interval bounds instead of standard errors and t-statistics. Set 'confint' to TRUE.

```r
sim_slopes(model, pred = x, modx = z, 
           confint = TRUE, ci.width = 0.95)
```

--------------------------------

### Margins Analysis with Delta Method Standard Errors

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_margins.md

Computes marginal effects using the delta method for standard error estimation, which is the default and fastest method.

```r
# Delta method (default, fast)
results <- sim_margins(fit, pred = hp, modx = wt, vce = "delta")
```

--------------------------------

### Using Robust Standard Errors

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/00-START-HERE.md

For analyses with heteroscedastic data, `sim_slopes` supports robust standard errors. Specify the type of robust SEs using the `robust` argument, such as `"HC3"` for heteroscedasticity-consistent standard errors.

```r
# Use robust SEs with messy data
sim_slopes(model, pred = x, modx = z, robust = "HC3")
```

--------------------------------

### Tidy Simple Margins Results

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Converts the results of a sim_margins analysis into a tidy data frame, similar in structure to tidy.sim_slopes.

```r
results <- sim_margins(model, pred = x, modx = z)
tidy_results <- tidy(results)
```

--------------------------------

### Combine Interaction Analysis and Plotting

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/INDEX.md

The probe_interaction function conveniently combines the functionality of sim_slopes and interact_plot into a single call. It simplifies the process of performing comprehensive interaction analysis and visualization.

```r
results <- probe_interaction(model, pred = x, modx = z)
```

--------------------------------

### Plot Simple Slopes

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/s3-methods.md

Generates a plot of simple slopes from a sim_slopes object. It may delegate to Johnson-Neyman plotting if available. For more control, consider using johnson_neyman() directly.

```r
results <- sim_slopes(model, pred = x, modx = z, johnson_neyman = TRUE)
plot(results)
```

--------------------------------

### Margins Analysis with Custom Confidence Intervals

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_margins.md

Displays confidence intervals instead of standard errors for marginal effects. Allows specifying the confidence interval width.

```r
# Show CIs instead of standard errors
results <- sim_margins(fit, pred = hp, modx = wt,
                       confint = TRUE, ci.width = 0.99)
results
```

--------------------------------

### Plot Interaction Effects with Confidence Intervals

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/00-START-HERE.md

Generates a plot to visualize interaction effects, including confidence intervals. This helps in understanding the nature and significance of the interaction.

```r
interact_plot(model, pred = x, modx = z, interval = TRUE)
```

--------------------------------

### sim_margins Function

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/api-reference-sim_margins.md

Conducts a simple margins analysis for understanding interaction effects in regression. It is similar to sim_slopes but uses the margins package approach with different standard error calculations. This function is suitable for analyzing two- and three-way interactions.

```APIDOC
## sim_margins

### Description
Conducts a simple margins analysis for understanding interaction effects in regression models. It is suitable for two- and three-way interactions and utilizes the `margins` package approach for standard error calculations.

### Function Signature
```r
sim_margins(
  model,
  pred,
  modx,
  mod2 = NULL,
  modx.values = NULL,
  mod2.values = NULL,
  data = NULL,
  cond.int = FALSE,
  vce = c("delta", "simulation", "bootstrap", "none"),
  iterations = 1000,
  digits = getOption("jtools-digits", default = 2),
  pvals = TRUE,
  confint = FALSE,
  ci.width = 0.95,
  cluster = NULL,
  modx.labels = NULL,
  mod2.labels = NULL,
  ...
)
```

### Parameters

#### Model Parameters

- **model** (regression object) - Required - Fitted model (lm, glm, svyglm, merMod, etc.). Must include the interaction.
- **pred** (symbol/string) - Required - Focal predictor variable. Evaluated with rlang (allows !! syntax). Must be numeric.
- **modx** (symbol/string) - Required - Moderator variable. Evaluated with rlang.
- **mod2** (symbol/string) - Optional - Optional second moderator for three-way interactions.

#### Value Specification

- **modx.values** (vector/character) - Optional - Values at which to compute margins. NULL (or "mean-plus-minus") uses mean ± 1 SD for numeric. "plus-minus", "terciles" also available. Or numeric vector.
- **mod2.values** (vector/character) - Optional - Values for second moderator. Same options as modx.values.

#### Data and Options

- **data** (data.frame) - Optional - Original data used to fit model. Improves centering.
- **cond.int** (logical) - Optional - Include conditional intercepts in output? Default is FALSE.

#### Standard Error and Output Control

- **vce** (character) - Optional - Method for standard error calculation: "delta" (default, fast), "simulation", "bootstrap", or "none".
- **iterations** (integer) - Optional - Number of iterations for "simulation" or "bootstrap" vce methods. Default is 1000.
- **digits** (integer) - Optional - Decimal places in output. Default is 2.
- **pvals** (logical) - Optional - Include p-values? Default is TRUE.
- **confint** (logical) - Optional - Show confidence intervals instead of standard errors? Default is FALSE.
- **ci.width** (numeric) - Optional - Confidence interval width (0-1). Ignored if confint=FALSE. Default is 0.95.
- **cluster** (character/vector) - Optional - Column name of cluster variable or vector for clustered standard errors.

#### Labels

- **modx.labels** (character vector) - Optional - Labels for moderator values. If NULL, values themselves used.
- **mod2.labels** (character vector) - Optional - Labels for second moderator values.

#### Additional Arguments

- **...** - Optional - Arguments passed to `margins::margins()`.
```

--------------------------------

### Check Object Class

Source: https://github.com/jacob-long/interactions/blob/master/_autodocs/types.md

Check the class of an object using the `class()` function. Use `inherits()` to specifically check if an object belongs to the 'sim_slopes' class.

```r
class(object)
inherits(object, "sim_slopes")
is_sim_slopes <- inherits(object, "sim_slopes")
```