-
-
Notifications
You must be signed in to change notification settings - Fork 94
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
11 changed files
with
601 additions
and
376 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,145 @@ | ||
|
|
||
| ```{r fig1_prep} | ||
| update_geom_defaults("point", aes(size = 3, color = "black", fill = "grey80", stroke = 1)) | ||
| theme_set( | ||
| theme_classic() + | ||
| theme( | ||
| axis.line = element_blank(), | ||
| panel.border = element_rect(colour = "black", fill = NA, linewidth = rel(2)), | ||
| legend.position = "bottom", | ||
| plot.title.position = "plot" | ||
| ) | ||
| ) | ||
| # Data -------------------------------------------------------------------- | ||
| data("women", package = "datasets") | ||
| # Convert units to metric | ||
| women$height <- women$height * 2.54 | ||
| women$weight <- women$weight / 2.205 | ||
| # Make some more data | ||
| set.seed(42) | ||
| women <- rbind( | ||
| women, | ||
| # duplicate with jitter | ||
| apply(women, 2, jitter, factor = 5), | ||
| # short and overweight | ||
| c(155, 73), | ||
| # just tall | ||
| c(230, 103) | ||
| ) | ||
| # Descriptives | ||
| ranges <- lapply(women, range) | ||
| M <- sapply(women[, 1:2], mean) | ||
| V <- cov(women[, 1:2]) | ||
| s <- sqrt(diag(V)) | ||
| # Plot constants | ||
| L <- list( | ||
| scale_shape_manual(NULL, labels = c("Not an Outlier", "Outlier"), values = c(20, 24)), | ||
| coord_cartesian(ranges[["height"]], ranges[["weight"]]), | ||
| labs(x = "Height [cm]", y = "Weight [kg]") | ||
| ) | ||
| # Univariate methods ------------------------------------------------------ | ||
| women[["univ_outlier"]] <- check_outliers(women, method = "zscore") | ||
| Z <- c(-3.290527, -2, -1, 0, 1, 2, 3.290527) | ||
| uni_thresholds <- data.frame( | ||
| X = M["height"] + s["height"] * Z, | ||
| Y = M["weight"] + s["weight"] * Z, | ||
| Z_lab = Z |> | ||
| format_value(digits = "signif3", | ||
| style_positive = "plus", | ||
| style_negative = "minus") |> | ||
| paste0(" SD") |> | ||
| replace(4, "Mean") | ||
| ) | ||
| p_uni <- ggplot(women, aes(height, weight)) + | ||
| geom_texthline(aes(yintercept = Y, label = Z_lab), data = uni_thresholds, | ||
| hjust = 0.8, halign = "center", | ||
| color = "grey60", size = 3) + | ||
| geom_point(aes(shape = univ_outlier)) + | ||
| L + | ||
| labs(title = "(A) Univariate") | ||
| # Multivariate methods -------------------------------------------- | ||
| women[["multiv_outlier"]] <- check_outliers(women[,1:2], method = "mahalanobis") | ||
| multi_thresholds <- qchisq(p = 0.001, df = 2, lower.tail = FALSE) | ||
| Ds <- c(1, 2, sqrt(multi_thresholds)) | ||
| ellipses <- lapply(setNames(nm = Ds), function(.D) { | ||
| coords <- ellipse( | ||
| center = M, | ||
| shape = V, | ||
| radius = .D, | ||
| segments = 100, | ||
| draw = FALSE | ||
| ) | ||
| coords <- as.data.frame(coords) | ||
| coords[["D"]] <- .D | ||
| coords | ||
| }) | ||
| ellipses <- do.call("rbind", ellipses) | ||
| colnames(ellipses)[1:2] <- colnames(women)[1:2] | ||
| ellipses[["D_lab"]] <- format_value(ellipses[["D"]], digits = "signif3") | ||
| p_multi <- | ||
| ggplot(women, aes(height, weight)) + | ||
| geom_point(aes(x = M["height"], y = M["weight"]), | ||
| color = "grey60", size = 4) + | ||
| geom_textcontour(aes(group = D_lab, label = D_lab), data = ellipses, | ||
| stat = "identity", hjust = 0.17, | ||
| color = "grey60", size = 3) + | ||
| geom_point(aes(shape = multiv_outlier)) + | ||
| L + | ||
| labs(title = "(B) Multivariate") | ||
| # Model-specific methods ------------------------------------------------ | ||
| model <- lm(weight ~ height, data = women) | ||
| women[["model_outlier"]] <- check_outliers(model, method = "cook") | ||
| women[["y_hat"]] <- fitted(model) | ||
| a <- coef(model)[1] | ||
| b <- coef(model)[2] | ||
| p_model <- | ||
| ggplot(women, aes(height, weight)) + | ||
| geom_abline(intercept = a, slope = b, color = "grey60") + | ||
| geom_segment(aes(xend = height, yend = y_hat), | ||
| color = "grey60") + | ||
| geom_point(aes(shape = model_outlier)) + | ||
| L + | ||
| labs(title = "(C) Model") | ||
| # Combine plots --------------------------------------------------- | ||
| no_y <- theme(axis.text.y = element_blank(), | ||
| axis.title.y = element_blank()) | ||
| fig1 <- p_uni + (p_multi + no_y) + (p_model + no_y) + | ||
| plot_layout(nrow = 1, guides = "collect") | ||
| ``` |
Large diffs are not rendered by default.
Oops, something went wrong.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Binary file not shown.
Oops, something went wrong.