diff --git a/tests/testthat/test-renderer-k-nearest-neighbour.R b/tests/testthat/test-renderer-k-nearest-neighbour.R
new file mode 100644
index 00000000..a15266c8
--- /dev/null
+++ b/tests/testthat/test-renderer-k-nearest-neighbour.R
@@ -0,0 +1,407 @@
+acontext("Proximity kNN Animation")
+
+library(dplyr)
+library(animint2)
+library(XML)
+
+# Helper Function: kNN Animint
+knn_animint <- function(train, test, class.vec, k = 30,
+                        interval = 1000,
+                        out.dir = "knn_animint",
+                        github_repo = NULL) {
+  # 1. Prepare Data
+  train.df <- as.data.frame(train)
+  colnames(train.df) <- c("x", "y")
+  train.df$class <- as.character(class.vec)
+  train.df$train.id <- seq_len(nrow(train.df))
+
+  test.df <- as.data.frame(test)
+  colnames(test.df) <- c("x", "y")
+  n.test <- nrow(test.df)
+
+  # k sequence
+  if (k <= 5) {
+    k.sequence <- 1:k
+  } else {
+    k.sequence <- seq(1, min(10, k), by = 1)
+    if (k >= 15) {
+      k.sequence <- c(k.sequence, seq(15, k, by = 5))
+    }
+    if (!k %in% k.sequence) k.sequence <- c(k.sequence, k)
+    k.sequence <- sort(unique(k.sequence))
+  }
+
+  # 2. Build Animation Frames
+  all.data.list <- list()
+  frame.id <- 1
+
+  for (test.idx in 1:n.test) {
+    test.point <- test.df[test.idx, ]
+
+    # Calculate distances
+    distances <- sqrt((train.df$x - test.point$x)^2 + (train.df$y - test.point$y)^2)
+    train.sorted <- train.df
+    train.sorted$distance <- distances
+    train.sorted <- train.sorted[order(distances), ]
+    train.sorted$rank <- seq_len(nrow(train.sorted))
+
+    for (k.val in k.sequence) {
+      train.sorted$is.knn <- train.sorted$rank <= k.val
+
+      # Votes
+      knn.subset <- train.sorted[train.sorted$is.knn, ]
+      votes <- knn.subset %>%
+        group_by(class) %>%
+        dplyr::summarise(votes = dplyr::n(), .groups = "drop")
+
+      predicted <- votes %>%
+        arrange(desc(votes)) %>%
+        slice(1) %>%
+        pull(class)
+
+      # Training points with aesthetic variables
+      train.frame <- train.sorted %>%
+        mutate(
+          frame.id = frame.id,
+          k = k.val,
+          test.idx = test.idx,
+          is.neighbor = is.knn,
+          # Visual properties based on neighbor status
+          point.alpha = ifelse(is.knn, 1.0, 0.3),
+          point.stroke = ifelse(is.knn, 1.2, 0.5)
+        )
+
+      # Lines to k-nearest neighbors
+      lines.frame <- train.sorted %>%
+        filter(is.knn) %>%
+        mutate(
+          x.start = test.point$x,
+          y.start = test.point$y,
+          x.end = x,
+          y.end = y,
+          frame.id = frame.id,
+          line.group = paste0("test", test.idx, "_train", train.id)
+        )
+
+      # Convex hull polygon
+      polygon.frame <- data.frame()
+      if (sum(train.sorted$is.knn) >= 3) {
+        knn.points <- train.sorted[train.sorted$is.knn, ]
+        hull.idx <- chull(knn.points$x, knn.points$y)
+        polygon.frame <- knn.points[hull.idx, c("x", "y")]
+        polygon.frame$frame.id <- frame.id
+        polygon.frame$poly.group <- paste0("test", test.idx, "_k", k.val)
+      }
+
+      # Circle at k-th neighbor
+      radius <- train.sorted$distance[k.val]
+      angles <- seq(0, 2 * pi, length.out = 100)
+      circle.frame <- data.frame(
+        x = test.point$x + radius * cos(angles),
+        y = test.point$y + radius * sin(angles),
+        frame.id = frame.id,
+        circle.group = paste0("test", test.idx, "_k", k.val)
+      )
+
+      # Current test point
+      test.current <- data.frame(
+        x = test.point$x,
+        y = test.point$y,
+        frame.id = frame.id
+      )
+
+      # Previously classified
+      prev.classified <- data.frame()
+      if (test.idx > 1) {
+        for (prev.idx in 1:(test.idx - 1)) {
+          prev.point <- test.df[prev.idx, ]
+          prev.dist <- sqrt((train.df$x - prev.point$x)^2 + (train.df$y - prev.point$y)^2)
+          prev.sorted <- train.df[order(prev.dist), ]
+          prev.knn <- prev.sorted[1:k, ]
+          prev.vote <- prev.knn %>%
+            group_by(class) %>%
+            dplyr::summarise(n = dplyr::n(), .groups = "drop") %>%
+            arrange(desc(n)) %>%
+            slice(1)
+
+          prev.classified <- rbind(prev.classified, data.frame(
+            x = prev.point$x,
+            y = prev.point$y,
+            class = prev.vote$class,
+            frame.id = frame.id
+          ))
+        }
+      }
+
+      # Future unclassified
+      future.points <- data.frame()
+      if (test.idx < n.test) {
+        future.points <- test.df[(test.idx + 1):n.test, ] %>%
+          mutate(frame.id = frame.id)
+      }
+
+      # Votes
+      votes.frame <- votes %>%
+        mutate(frame.id = frame.id, k = k.val)
+
+      # Title
+      title.text <- sprintf(
+        "Test Point %d/%d | k = %d | Predicted: %s",
+        test.idx, n.test, k.val, predicted
+      )
+
+      all.data.list[[frame.id]] <- list(
+        training = train.frame,
+        lines = lines.frame,
+        polygon = polygon.frame,
+        circle = circle.frame,
+        test_current = test.current,
+        test_prev = prev.classified,
+        test_future = future.points,
+        votes = votes.frame,
+        title = data.frame(
+          x = mean(range(train.df$x)),
+          y = max(train.df$y) + 0.8,
+          label = title.text,
+          frame.id = frame.id
+        )
+      )
+
+      frame.id <- frame.id + 1
+    }
+  }
+
+  # 3. Combine Data
+
+  extract_element <- function(element.name) {
+    bind_rows(lapply(all.data.list, function(x) x[[element.name]]))
+  }
+
+  training.all <- extract_element("training")
+  lines.all <- extract_element("lines")
+  polygon.all <- extract_element("polygon")
+  circle.all <- extract_element("circle")
+  test.current.all <- extract_element("test_current")
+  test.prev.all <- extract_element("test_prev")
+  test.future.all <- extract_element("test_future")
+  votes.all <- extract_element("votes")
+  title.all <- extract_element("title")
+
+  # 4. Color Scheme
+  unique.classes <- unique(class.vec)
+  n.classes <- length(unique.classes)
+
+  # Constants mapping
+  class.colors <- c("#4472C4", "#70AD47", "#FFC000", "#C00000")[1:n.classes]
+  names(class.colors) <- unique.classes
+
+  class.shapes <- c(1, 2, 0, 5)[1:n.classes]
+  names(class.shapes) <- unique.classes
+
+  # Calculate proper axis ranges
+  x.range <- range(c(train.df$x, test.df$x))
+  y.range <- range(c(train.df$y, test.df$y))
+  x.padding <- diff(x.range) * 0.1
+  y.padding <- diff(y.range) * 0.1
+
+  # 5. Create Plots
+
+  # Main Feature Space Plot
+  p.feature <- ggplot() +
+    geom_polygon(
+      data = polygon.all,
+      aes(x = x, y = y, group = poly.group),
+      fill = "#FFFFCC",
+      alpha = 0.4,
+      showSelected = "frame.id"
+    ) +
+    geom_path(
+      data = circle.all,
+      aes(x = x, y = y, group = circle.group),
+      color = "gray40",
+      linetype = "dashed",
+      size = 1,
+      showSelected = "frame.id"
+    ) +
+    geom_segment(
+      data = lines.all,
+      aes(x = x.start, y = y.start, xend = x.end, yend = y.end, group = line.group),
+      color = "gray60",
+      linetype = "dotted",
+      size = 0.8,
+      alpha = 0.6,
+      showSelected = "frame.id"
+    ) +
+    # Combined Training Points (Neighbors and Non-neighbors)
+    # Using aesthetic mapping for alpha and stroke to differentiate
+    geom_point(
+      data = training.all,
+      aes(
+        x = x, y = y, fill = class,
+        alpha = point.alpha, stroke = I(point.stroke)
+      ),
+      shape = 21,
+      color = "gray50",
+      size = 2.5,
+      showSelected = "frame.id"
+    ) +
+    scale_alpha_identity() + # Use literal values from data for alpha
+    geom_point(
+      data = test.prev.all,
+      aes(x = x, y = y, fill = class),
+      shape = 21,
+      color = "black",
+      size = 2.5,
+      alpha = 1.0,
+      stroke = 1.5,
+      showSelected = "frame.id"
+    ) +
+    geom_text(
+      data = test.future.all,
+      aes(x = x, y = y),
+      label = "?",
+      color = "#8B4513",
+      size = 8,
+      fontface = "bold",
+      vjust = 0,
+      showSelected = "frame.id"
+    ) +
+    geom_point(
+      data = test.current.all,
+      aes(x = x, y = y),
+      shape = 21,
+      fill = "#FF0000",
+      color = "black",
+      size = 3,
+      alpha = 1.0,
+      showSelected = "frame.id"
+    ) +
+    geom_point(
+      data = test.current.all,
+      aes(x = x, y = y),
+      shape = 21,
+      fill = "transparent",
+      color = "#FF0000",
+      size = 4,
+      stroke = 2,
+      showSelected = "frame.id"
+    ) +
+    geom_text(
+      data = title.all,
+      aes(x = x, y = y, label = label),
+      size = 5,
+      fontface = "bold",
+      vjust = 0,
+      showSelected = "frame.id"
+    ) +
+    scale_fill_manual(values = class.colors, name = "Class") +
+    theme_animint(width = 900, height = 700) +
+    scale_x_continuous(
+      breaks = pretty(x.range, n = 10),
+      limits = c(x.range[1] - x.padding, x.range[2] + x.padding)
+    ) +
+    scale_y_continuous(
+      breaks = pretty(y.range, n = 10),
+      limits = c(y.range[1] - y.padding, y.range[2] + y.padding)
+    ) +
+    coord_fixed() +
+    labs(x = "x1", y = "x2") +
+    theme_bw() +
+    theme(
+      legend.position = "top",
+      panel.grid.major = element_line(color = "gray85"),
+      panel.grid.minor = element_line(color = "gray92"),
+      axis.text = element_text(size = 10),
+      axis.title = element_text(size = 12)
+    )
+
+  # Vote Chart
+  p.votes <- ggplot() +
+    geom_bar(
+      data = votes.all,
+      aes(x = class, y = votes, fill = class),
+      stat = "identity",
+      position = "identity",
+      width = 0.6,
+      showSelected = "frame.id"
+    ) +
+    scale_fill_manual(values = class.colors) +
+    scale_y_continuous(
+      breaks = seq(0, k, by = 5),
+      limits = c(0, k)
+    ) +
+    labs(
+      title = "Classification Votes",
+      x = "Class",
+      y = "Votes"
+    ) +
+    theme_bw() +
+    theme(
+      legend.position = "none",
+      plot.title = element_text(hjust = 0.5, face = "bold"),
+      axis.text = element_text(size = 10),
+      axis.title = element_text(size = 11)
+    ) +
+    theme_animint(width = 400, height = 500)
+
+  # 6. Create animint
+  animint(
+    featureSpace = p.feature,
+    voteChart = p.votes,
+    title = "k-Nearest Neighbors Classification",
+    time = list(
+      variable = "frame.id",
+      ms = interval
+    ),
+    first = list(frame.id = 1)
+  )
+}
+
+test_that("kNN animation renders correctly", {
+  set.seed(42)
+
+  train.data <- matrix(c(rnorm(40, mean = -1), rnorm(40, mean = 1)),
+    ncol = 2, byrow = TRUE
+  )
+  test.data <- matrix(rnorm(4, mean = 0, sd = 1.2), ncol = 2)
+  classes <- rep(c("Class A", "Class B"), each = 20)
+
+  viz <- knn_animint(
+    train = train.data,
+    test = test.data,
+    class.vec = classes,
+    k = 10,
+    interval = 500
+  )
+
+  info <- animint2HTML(viz)
+
+  # 1. Check title exists and is correct
+  expect_true(nzchar(saveXML(info$html)) > 0)
+
+  # Check <title> tag for the main title
+  page.title <- xmlValue(getNodeSet(info$html, "//title")[[1]])
+  expect_equal(page.title, "k-Nearest Neighbors Classification")
+
+  # 2. Check for plots presence using SVGs with plot_* IDs
+  feature.node <- getNodeSet(info$html, '//svg[@id="plot_featureSpace"]')
+  vote.node <- getNodeSet(info$html, '//svg[@id="plot_voteChart"]')
+
+  expect_equal(length(feature.node), 1)
+  expect_equal(length(vote.node), 1)
+
+  # 3. Check for specific text appearing in the SVG
+  svg.text <- getNodeSet(info$html, "//text")
+  all.text <- sapply(svg.text, xmlValue)
+  expect_true(any(grepl("Classification Votes", all.text)))
+
+  # 4. Check visual elements
+  circles <- getNodeSet(info$html, "//circle")
+  expect_true(length(circles) > 0)
+
+  paths <- getNodeSet(info$html, "//path")
+  expect_true(length(paths) > 0)
+
+  rects <- getNodeSet(info$html, "//rect")
+  expect_true(length(rects) > 0)
+})