diff --git a/R/affinity.R b/R/affinity.R
index 284d097d..10844495 100644
--- a/R/affinity.R
+++ b/R/affinity.R
@@ -11,7 +11,8 @@ fuzzy_set_union <- function(X, set_op_mix_ratio = 1) {
   }
   else {
     Matrix::drop0(
-      set_op_mix_ratio * (X + Matrix::t(X) - XX) + (1 - set_op_mix_ratio) * XX)
+      set_op_mix_ratio * (X + Matrix::t(X) - XX) + (1 - set_op_mix_ratio) * XX
+    )
   }
 }
 
@@ -19,8 +20,10 @@ fuzzy_set_union <- function(X, set_op_mix_ratio = 1) {
 # or not
 smooth_knn <- function(nn,
                        local_connectivity = 1.0, bandwidth = 1.0,
-                       n_threads = max(1, 
-                                       RcppParallel::defaultNumThreads() / 2),
+                       n_threads = max(
+                         1,
+                         RcppParallel::defaultNumThreads() / 2
+                       ),
                        grain_size = 1,
                        verbose = FALSE) {
   tsmessage(
@@ -52,9 +55,11 @@ smooth_knn <- function(nn,
 fuzzy_simplicial_set <- function(nn,
                                  set_op_mix_ratio = 1.0,
                                  local_connectivity = 1.0, bandwidth = 1.0,
-                                 n_threads = 
-                                   max(1, 
-                                       RcppParallel::defaultNumThreads() / 2),
+                                 n_threads =
+                                   max(
+                                     1,
+                                     RcppParallel::defaultNumThreads() / 2
+                                   ),
                                  grain_size = 1,
                                  verbose = FALSE) {
   affinity_matrix <- smooth_knn(nn,
@@ -77,9 +82,11 @@ symmetrize <- function(P) {
 }
 
 perplexity_similarities <- function(nn, perplexity = NULL,
-                                    n_threads = 
-                                      max(1, 
-                                          RcppParallel::defaultNumThreads() / 2),
+                                    n_threads =
+                                      max(
+                                        1,
+                                        RcppParallel::defaultNumThreads() / 2
+                                      ),
                                     grain_size = 1,
                                     kernel = "gauss",
                                     verbose = FALSE) {
@@ -118,8 +125,9 @@ perplexity_similarities <- function(nn, perplexity = NULL,
 # edge has a weight of val (scalar or vector)
 # return a sparse matrix with dimensions of nrow(nn_idx) x max_nbr_id
 nn_to_sparse <- function(nn_idx, val = 1, byrow = FALSE, self_nbr = FALSE,
-                         max_nbr_id = ifelse(self_nbr, 
-                                             nrow(nn_idx), max(nn_idx))) {
+                         max_nbr_id = ifelse(self_nbr,
+                           nrow(nn_idx), max(nn_idx)
+                         )) {
   nd <- nrow(nn_idx)
   k <- ncol(nn_idx)
 
diff --git a/R/init.R b/R/init.R
index f37ec884..5d27e131 100644
--- a/R/init.R
+++ b/R/init.R
@@ -30,8 +30,10 @@ laplacian_eigenmap <- function(A, ndim = 2, verbose = FALSE) {
   )
 
   if (is.null(eig_res) || ncol(eig_res$vectors) < ndim + 1) {
-    message("Laplacian Eigenmap failed to converge, ", 
-            "using random initialization instead")
+    message(
+      "Laplacian Eigenmap failed to converge, ",
+      "using random initialization instead"
+    )
     return(rand_init(nrow(A), ndim))
   }
   vecs <- eig_res$vectors[, 2:(ndim + 1)]
@@ -66,8 +68,10 @@ normalized_laplacian_init <- function(A, ndim = 2, verbose = FALSE) {
     }
   )
   if (is.null(res) || ncol(res$vectors) < ndim) {
-    message("Spectral initialization failed to converge, ",
-            "using random initialization instead")
+    message(
+      "Spectral initialization failed to converge, ",
+      "using random initialization instead"
+    )
     return(rand_init(n, ndim))
   }
   vec_indices <- rev(order(res$values, decreasing = TRUE)[1:ndim])
diff --git a/R/neighbors.R b/R/neighbors.R
index a8149e2f..80a7b234 100644
--- a/R/neighbors.R
+++ b/R/neighbors.R
@@ -225,8 +225,10 @@ sparse_nn <- function(X, k, include_self = TRUE) {
     is_nonzero <- dists != 0
     dist_nonzero <- dists[is_nonzero]
     if (length(dist_nonzero) < k) {
-      stop("Row ", i, " of distance matrix has only ", length(dist_nonzero),
-           " defined distances")
+      stop(
+        "Row ", i, " of distance matrix has only ", length(dist_nonzero),
+        " defined distances"
+      )
     }
 
     k_order <- order(dist_nonzero)[1:k]
diff --git a/R/supervised.R b/R/supervised.R
index 17bdaa94..ff5dbe7d 100644
--- a/R/supervised.R
+++ b/R/supervised.R
@@ -10,8 +10,8 @@
 # Return The resulting intersected fuzzy simplicial set.
 categorical_simplicial_set_intersection <- function(
                                                     simplicial_set, target,
-                                                    unknown_dist = 1.0, 
-                                                    far_dist = 5.0, 
+                                                    unknown_dist = 1.0,
+                                                    far_dist = 5.0,
                                                     verbose = FALSE) {
 
   # Convert to dgTMatrix to get to the j indices
@@ -113,11 +113,11 @@ general_sset_intersection <- function(indptr1,
 
     if (left_val > left_min || right_val > right_min) {
       if (mix_weight < 0.5) {
-        result_val[idx] <- left_val * 
+        result_val[idx] <- left_val *
           right_val^(mix_weight / (1.0 - mix_weight))
       }
       else {
-        result_val[idx] <- right_val * 
+        result_val[idx] <- right_val *
           left_val^(((1.0 - mix_weight) / mix_weight))
       }
     }
diff --git a/R/transform.R b/R/transform.R
index b940edeb..ac121827 100644
--- a/R/transform.R
+++ b/R/transform.R
@@ -42,7 +42,7 @@ umap_transform <- function(X, model,
                            init_weighted = TRUE,
                            search_k = NULL,
                            n_epochs = NULL,
-                           n_threads = 
+                           n_threads =
                              max(1, RcppParallel::defaultNumThreads() / 2),
                            grain_size = 1,
                            verbose = FALSE) {
@@ -181,7 +181,7 @@ umap_transform <- function(X, model,
 }
 
 init_new_embedding <- function(train_embedding, nn, graph, weighted = TRUE,
-                               n_threads = 
+                               n_threads =
                                  max(1, RcppParallel::defaultNumThreads() / 2),
                                grain_size = 1, verbose = FALSE) {
   parallelize <- n_threads > 0
diff --git a/R/uwot.R b/R/uwot.R
index a1762071..f58fd13d 100644
--- a/R/uwot.R
+++ b/R/uwot.R
@@ -365,7 +365,7 @@ tumap <- function(X, n_neighbors = 15, n_components = 2, metric = "euclidean",
   uwot(
     X = X, n_neighbors = n_neighbors, n_components = n_components,
     metric = metric,
-    n_epochs = n_epochs, alpha = alpha, scale = scale, init = init, 
+    n_epochs = n_epochs, alpha = alpha, scale = scale, init = init,
     spread = NULL, min_dist = NULL, set_op_mix_ratio = set_op_mix_ratio,
     local_connectivity = local_connectivity, bandwidth = bandwidth,
     gamma = gamma, negative_sample_rate = negative_sample_rate,
@@ -536,7 +536,7 @@ lvish <- function(X, perplexity = 50, n_neighbors = perplexity * 3,
   uwot(X,
     n_neighbors = n_neighbors, n_components = n_components,
     metric = metric,
-    n_epochs = n_epochs, alpha = alpha, scale = scale, init = init, 
+    n_epochs = n_epochs, alpha = alpha, scale = scale, init = init,
     gamma = gamma, negative_sample_rate = negative_sample_rate,
     nn_method = nn_method, n_trees = n_trees, search_k = search_k,
     method = "largevis", perplexity = perplexity,
@@ -548,7 +548,7 @@ lvish <- function(X, perplexity = 50, n_neighbors = perplexity * 3,
 # Function that does all the real work
 uwot <- function(X, n_neighbors = 15, n_components = 2, metric = "euclidean",
                  n_epochs = NULL,
-                 alpha = 1, scale = FALSE, init = "spectral", 
+                 alpha = 1, scale = FALSE, init = "spectral",
                  spread = 1, min_dist = 0.01,
                  set_op_mix_ratio = 1.0, local_connectivity = 1.0,
                  bandwidth = 1.0, gamma = 1.0,
@@ -556,7 +556,7 @@ uwot <- function(X, n_neighbors = 15, n_components = 2, metric = "euclidean",
                  nn_method = NULL, n_trees = 50,
                  search_k = 2 * n_neighbors * n_trees,
                  method = "umap", perplexity = 50, approx_pow = FALSE,
-                 y = NULL, target_n_neighbors = n_neighbors, 
+                 y = NULL, target_n_neighbors = n_neighbors,
                  target_weight = 0.5,
                  n_threads = max(1, RcppParallel::defaultNumThreads() / 2),
                  kernel = "gauss",
@@ -651,8 +651,10 @@ uwot <- function(X, n_neighbors = 15, n_components = 2, metric = "euclidean",
   }
   nn_method <- match.arg(tolower(nn_method), c("annoy", "fnn"))
   if (nn_method == "fnn" && metric != "euclidean") {
-    stop("nn_method = 'FNN' is only compatible with distance metric ",
-         "'euclidean'")
+    stop(
+      "nn_method = 'FNN' is only compatible with distance metric ",
+      "'euclidean'"
+    )
   }
   if (nn_method == "fnn" && ret_model) {
     stop("nn_method = 'FNN' is incompatible with ret_model = TRUE")
@@ -949,8 +951,10 @@ scale_input <- function(X, scale_type, ret_model = FALSE, verbose = FALSE) {
     scale_type <- ifelse(scale_type, "scale", "none")
   }
 
-  scale_type <- match.arg(tolower(scale_type), 
-                          c("none", "scale", "range", "maxabs"))
+  scale_type <- match.arg(
+    tolower(scale_type),
+    c("none", "scale", "range", "maxabs")
+  )
   switch(scale_type,
     range = {
       tsmessage("Range scaling X")
diff --git a/tests/testthat/helper_fuzzy_sets.R b/tests/testthat/helper_fuzzy_sets.R
index 13512bbf..6c20bb53 100644
--- a/tests/testthat/helper_fuzzy_sets.R
+++ b/tests/testthat/helper_fuzzy_sets.R
@@ -3,13 +3,17 @@
 
 # Asymmetric fuzzy set data
 V_asymm <- sparseMatrix(
-  i = c(5, 6, 8, 3, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 8, 3, 5, 1,
-        5, 6, 7, 4, 1, 2, 4, 8),
-  j = c(1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 7, 7, 8,
-        8, 8, 8, 9, 10, 10, 10, 10),
+  i = c(
+    5, 6, 8, 3, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 8, 3, 5, 1,
+    5, 6, 7, 4, 1, 2, 4, 8
+  ),
+  j = c(
+    1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 7, 7, 8,
+    8, 8, 8, 9, 10, 10, 10, 10
+  ),
   x = c(
     1, 1, 1, 0.328, 0.4252, 1, 0.5386, 1, 1, 0.5748, 0.5, 0.4614, 1, 0.6463,
-    1, 0.5, 1, 1, 0.6894, 0.672, 0.2807, 0.8381, 0.7193, 5.129e-10, 0.3538, 
+    1, 0.5, 1, 1, 0.6894, 0.672, 0.2807, 0.8381, 0.7193, 5.129e-10, 0.3538,
     0.5447, 0.1619, 1, 0.4553, 0.3106
   )
 )
@@ -25,9 +29,9 @@ V_union <- sparseMatrix(
     7, 7, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 10, 10, 10, 10, 10
   ),
   x = c(
-    1, 1, 1, 0.1619, 0.6899, 0.4614, 0.4252, 1, 0.6899, 1, 1, 0.5748, 0.5, 
+    1, 1, 1, 0.1619, 0.6899, 0.4614, 0.4252, 1, 0.6899, 1, 1, 0.5748, 0.5,
     0.4614, 1, 0.6463, 1, 0.7277, 1, 1, 0.2807, 0.9128, 1, 1, 5.129e-10, 1,
-    0.6463, 0.2807, 0.3538, 1, 0.9128, 5.129e-10, 0.3538, 0.3106, 0.4252, 
+    0.6463, 0.2807, 0.3538, 1, 0.9128, 5.129e-10, 0.3538, 0.3106, 0.4252,
     0.5748, 1, 0.1619, 1, 0.5, 0.7277, 0.3106
   )
 )
@@ -43,10 +47,10 @@ V_mix <- sparseMatrix(
     7, 7, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 10, 10, 10, 10, 10
   ),
   x = c(
-    1, 0.5, 0.919, 0.08095, 0.4333, 0.2307, 0.2126, 1, 0.4333, 1, 0.836, 
-    0.2874, 0.25, 0.2307, 1, 0.3231, 0.7723, 0.4777, 1, 0.5, 0.1404, 0.7043, 
-    0.5, 0.5, 2.564e-10, 0.836, 0.3231, 0.1404, 0.1769, 0.919, 0.7043, 
-    2.564e-10, 0.1769, 0.1553, 0.2126, 0.2874, 0.7723, 0.08095, 1, 0.25, 
+    1, 0.5, 0.919, 0.08095, 0.4333, 0.2307, 0.2126, 1, 0.4333, 1, 0.836,
+    0.2874, 0.25, 0.2307, 1, 0.3231, 0.7723, 0.4777, 1, 0.5, 0.1404, 0.7043,
+    0.5, 0.5, 2.564e-10, 0.836, 0.3231, 0.1404, 0.1769, 0.919, 0.7043,
+    2.564e-10, 0.1769, 0.1553, 0.2126, 0.2874, 0.7723, 0.08095, 1, 0.25,
     0.4777, 0.1553
   )
 )
@@ -63,10 +67,14 @@ V_intersect <- sparseMatrix(
 
 # asymm with local connectivity = 1.5
 V_asymm_local <- sparseMatrix(
-  i = c(5, 6, 8, 3, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 8, 3, 5, 1,
-        5, 6, 7, 4, 1, 2, 4, 8),
-  j = c(1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 7, 7, 8, 8,
-        8, 8, 9, 10, 10, 10, 10),
+  i = c(
+    5, 6, 8, 3, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 8, 3, 5, 1,
+    5, 6, 7, 4, 1, 2, 4, 8
+  ),
+  j = c(
+    1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 7, 7, 8, 8,
+    8, 8, 9, 10, 10, 10, 10
+  ),
   x = c(
     1, 1, 1, 0.2559, 0.3748, 1, 0.5698, 1, 1, 0.6252, 0.5, 0.4302, 1, 0.7157,
     1, 0.5, 1, 1, 0.7622, 0.7441, 0.2084, 0.8925, 0.7916, 5.129e-10, 0.2843,
@@ -76,17 +84,17 @@ V_asymm_local <- sparseMatrix(
 
 V_union_local <- sparseMatrix(
   i = c(
-    5, 6, 8, 10, 3, 4, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 7, 8, 1, 
+    5, 6, 8, 10, 3, 4, 9, 10, 2, 4, 7, 9, 10, 2, 3, 7, 9, 10, 1, 6, 7, 8, 1,
     5, 8, 3, 4, 5, 8, 1, 5, 6, 7, 10, 2, 3, 4, 1, 2, 3, 4, 8
   ),
   j = c(
-    1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 5, 6, 6, 
+    1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 5, 6, 6,
     6, 7, 7, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 10, 10, 10, 10, 10
   ),
   x = c(
-    1, 1, 1, 0.1075, 0.6799, 0.4302, 0.3748, 1, 0.6799, 1, 1, 0.6252, 0.5, 
-    0.4302, 1, 0.7157, 1, 0.7102, 1, 1, 0.2084, 0.9504, 1, 1, 5.129e-10, 1, 
-    0.7157, 0.2084, 0.2843, 1, 0.9504, 5.129e-10, 0.2843, 0.2378, 0.3748, 
+    1, 1, 1, 0.1075, 0.6799, 0.4302, 0.3748, 1, 0.6799, 1, 1, 0.6252, 0.5,
+    0.4302, 1, 0.7157, 1, 0.7102, 1, 1, 0.2084, 0.9504, 1, 1, 5.129e-10, 1,
+    0.7157, 0.2084, 0.2843, 1, 0.9504, 5.129e-10, 0.2843, 0.2378, 0.3748,
     0.6252, 1, 0.1075, 1, 0.5, 0.7102, 0.2378
   )
 )
@@ -101,9 +109,9 @@ V_union_bandwidth <- sparseMatrix(
     7, 7, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 10, 10, 10, 10, 10
   ),
   x = c(
-    1, 1, 1, 0.02621, 0.3664, 0.2129, 0.1808, 1, 0.3664, 1, 1, 0.3304, 0.25, 
-    0.2129, 1, 0.4176, 1, 0.4055, 1, 1, 0.07881, 0.7467, 1, 1, 2.63e-19, 1, 
-    0.4176, 0.07881, 0.1251, 1, 0.7467, 2.63e-19, 0.1251, 0.09647, 0.1808, 
+    1, 1, 1, 0.02621, 0.3664, 0.2129, 0.1808, 1, 0.3664, 1, 1, 0.3304, 0.25,
+    0.2129, 1, 0.4176, 1, 0.4055, 1, 1, 0.07881, 0.7467, 1, 1, 2.63e-19, 1,
+    0.4176, 0.07881, 0.1251, 1, 0.7467, 2.63e-19, 0.1251, 0.09647, 0.1808,
     0.3304, 1, 0.02621, 1, 0.25, 0.4055, 0.09647
   )
 )
@@ -111,6 +119,8 @@ V_union_bandwidth <- sparseMatrix(
 V_intersect_local_bandwidth <- sparseMatrix(
   i = c(5, 8, 3, 10, 2, 4, 7, 3, 9, 10, 1, 8, 3, 1, 5, 4, 2, 4),
   j = c(1, 1, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 7, 8, 8, 9, 10, 10),
-  x = c(1, 0.7966, 0.02126, 1, 0.02126, 1, 0.5536, 1, 0.336, 0.04417, 1, 0.364,
-        0.5536, 0.7966, 0.364, 0.336, 1, 0.04417)
+  x = c(
+    1, 0.7966, 0.02126, 1, 0.02126, 1, 0.5536, 1, 0.336, 0.04417, 1, 0.364,
+    0.5536, 0.7966, 0.364, 0.336, 1, 0.04417
+  )
 )
diff --git a/tests/testthat/test_knn_aff.R b/tests/testthat/test_knn_aff.R
index 6b35fe72..358bb0c3 100644
--- a/tests/testthat/test_knn_aff.R
+++ b/tests/testthat/test_knn_aff.R
@@ -43,5 +43,7 @@ expected_sym_nn_graph[10, c(1, 2, 3, 4, 8)] <- c(o6, o3, o6, o3, o6)
 
 expect_equal(sum(res), 10)
 expect_true(Matrix::isSymmetric(res))
-expect_equal(as.matrix(res), expected_sym_nn_graph, check.attributes = FALSE, 
-             tol = 1e-7)
+expect_equal(as.matrix(res), expected_sym_nn_graph,
+  check.attributes = FALSE,
+  tol = 1e-7
+)
diff --git a/tests/testthat/test_neighbors.R b/tests/testthat/test_neighbors.R
index 01a7a703..0b017c6b 100644
--- a/tests/testthat/test_neighbors.R
+++ b/tests/testthat/test_neighbors.R
@@ -115,8 +115,10 @@ res <- annoy_nn(ui10, k = 4, include_self = TRUE, n_threads = 0)
 expect_equal(res$idx, self_nn_index4, check.attributes = FALSE)
 expect_equal(res$dist, self_nn_dist4, check.attributes = FALSE, tol = 1e-6)
 
-res <- annoy_nn(ui10, k = 4, include_self = FALSE, n_threads = 0, 
-                ret_index = TRUE)
+res <- annoy_nn(ui10,
+  k = 4, include_self = FALSE, n_threads = 0,
+  ret_index = TRUE
+)
 expect_equal(res$idx, nn_index4, check.attributes = FALSE)
 expect_equal(res$dist, nn_dist4, check.attributes = FALSE, tol = 1e-6)
 expect_true(!is.null(res$index))
@@ -126,8 +128,10 @@ res <- annoy_nn(ui10, k = 4, include_self = TRUE, n_threads = 1)
 expect_equal(res$idx, self_nn_index4, check.attributes = FALSE)
 expect_equal(res$dist, self_nn_dist4, check.attributes = FALSE, tol = 1e-6)
 
-res <- annoy_nn(ui10, k = 4, include_self = FALSE, n_threads = 1, 
-                ret_index = TRUE)
+res <- annoy_nn(ui10,
+  k = 4, include_self = FALSE, n_threads = 1,
+  ret_index = TRUE
+)
 expect_equal(res$idx, nn_index4, check.attributes = FALSE)
 expect_equal(res$dist, nn_dist4, check.attributes = FALSE, tol = 1e-6)
 expect_true(!is.null(res$index))
diff --git a/tests/testthat/test_output.R b/tests/testthat/test_output.R
index 50ce414f..434ae096 100644
--- a/tests/testthat/test_output.R
+++ b/tests/testthat/test_output.R
@@ -29,20 +29,20 @@ expect_ok_matrix(res)
 
 # UMAP and cosine metric n_threads = 1 issue #5
 res <- umap(iris10,
-             n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "cosine",
-             init = "spectral", verbose = FALSE, n_threads = 1
+  n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "cosine",
+  init = "spectral", verbose = FALSE, n_threads = 1
 )
 expect_ok_matrix(res)
 
 # metric = Manhattan
 res <- umap(iris10,
-            n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "manhattan",
-            init = "rand", verbose = FALSE, n_threads = 0
+  n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "manhattan",
+  init = "rand", verbose = FALSE, n_threads = 0
 )
 expect_ok_matrix(res)
 res <- umap(iris10,
-            n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "manhattan",
-            init = "spca", verbose = FALSE, n_threads = 1
+  n_neighbors = 4, n_epochs = 2, alpha = 0.5, metric = "manhattan",
+  init = "spca", verbose = FALSE, n_threads = 1
 )
 expect_ok_matrix(res)
 
diff --git a/tests/testthat/test_smooth_knn_dists.R b/tests/testthat/test_smooth_knn_dists.R
index ffd831cd..e90dc320 100644
--- a/tests/testthat/test_smooth_knn_dists.R
+++ b/tests/testthat/test_smooth_knn_dists.R
@@ -50,16 +50,20 @@ expect_equal(res$P, V_asymm, tol = 1e-4)
 expect_equal(fuzzy_set_union(res$P), V_union, tol = 1e-4)
 
 # mix intersection with union
-expect_equal(fuzzy_set_union(res$P, set_op_mix_ratio = 0.5), V_mix, 
-             tol = 1e-4)
+expect_equal(fuzzy_set_union(res$P, set_op_mix_ratio = 0.5), V_mix,
+  tol = 1e-4
+)
 
 # intersection
-expect_equal(fuzzy_set_union(res$P, set_op_mix_ratio = 0), V_intersect, 
-             tol = 1e-4)
+expect_equal(fuzzy_set_union(res$P, set_op_mix_ratio = 0), V_intersect,
+  tol = 1e-4
+)
 
 # local connectivity
-res <- smooth_knn_distances(nn_4$dist, nn_4$idx, local_connectivity = 1.5, 
-                            ret_extra = TRUE)
+res <- smooth_knn_distances(nn_4$dist, nn_4$idx,
+  local_connectivity = 1.5,
+  ret_extra = TRUE
+)
 res$P <- nn_to_sparse(nn_4$idx, as.vector(res$P), self_nbr = TRUE)
 expect_equal(res$P, V_asymm_local, tol = 1e-4)
 
@@ -69,16 +73,18 @@ res_cpp_conn1 <- smooth_knn_distances_parallel(nn_4$dist, nn_4$idx,
   bandwidth = 1.0, tol = 1e-5, min_k_dist_scale = 1e-3,
   parallelize = FALSE, verbose = FALSE
 )
-expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1), 
-                          self_nbr = TRUE), V_asymm, tol = 1e-4)
+expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1),
+  self_nbr = TRUE
+), V_asymm, tol = 1e-4)
 
 res_cpp_conn1.5 <- smooth_knn_distances_parallel(nn_4$dist, nn_4$idx,
   n_iter = 64, local_connectivity = 1.5,
   bandwidth = 1.0, tol = 1e-5, min_k_dist_scale = 1e-3,
   parallelize = FALSE, verbose = FALSE
 )
-expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1.5), 
-                          self_nbr = TRUE), V_asymm_local, tol = 1e-4)
+expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1.5),
+  self_nbr = TRUE
+), V_asymm_local, tol = 1e-4)
 
 
 RcppParallel::setThreadOptions(numThreads = 1)
@@ -87,23 +93,27 @@ res_cpp_conn1 <- smooth_knn_distances_parallel(nn_4$dist, nn_4$idx,
   bandwidth = 1.0, tol = 1e-5, min_k_dist_scale = 1e-3,
   grain_size = 1, verbose = FALSE
 )
-expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1), 
-                          self_nbr = TRUE), V_asymm, tol = 1e-4)
+expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1),
+  self_nbr = TRUE
+), V_asymm, tol = 1e-4)
 
 res_cpp_conn1.5 <- smooth_knn_distances_parallel(nn_4$dist, nn_4$idx,
   n_iter = 64, local_connectivity = 1.5,
   bandwidth = 1.0, tol = 1e-5, min_k_dist_scale = 1e-3,
   grain_size = 1, verbose = FALSE
 )
-expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1.5), 
-                          self_nbr = TRUE), V_asymm_local, tol = 1e-4)
+expect_equal(nn_to_sparse(nn_4$idx, as.vector(res_cpp_conn1.5),
+  self_nbr = TRUE
+), V_asymm_local, tol = 1e-4)
 
 
 # Test cross-distances
 V_asymm_local_cross <- V_asymm_local
 diag(V_asymm_local_cross) <- 1
-V_asymm_local_cross <- cbind(V_asymm_local_cross, 
-                             matrix(0, nrow = 10, ncol = 2))
+V_asymm_local_cross <- cbind(
+  V_asymm_local_cross,
+  matrix(0, nrow = 10, ncol = 2)
+)
 
 res_cpp_conn1.5_cross <- smooth_knn_distances_parallel(nn_4$dist, nn_4$idx,
   n_iter = 64, local_connectivity = 1.5,
diff --git a/tests/testthat/test_transform.R b/tests/testthat/test_transform.R
index 1c5844ae..c80a5f8d 100644
--- a/tests/testthat/test_transform.R
+++ b/tests/testthat/test_transform.R
@@ -88,8 +88,10 @@ expect_equal(iris10_scale, iris10_strans, check.attributes = FALSE)
 
 iris10_zv_col <- iris10
 iris10_zv_col[, 3] <- 10
-iris10zvc_scale <- scale_input(iris10_zv_col, scale_type = "scale", 
-                               ret_model = TRUE)
+iris10zvc_scale <- scale_input(iris10_zv_col,
+  scale_type = "scale",
+  ret_model = TRUE
+)
 # scale the original iris10 here on purpose to check that full-variance column
 # is correctly removed
 iris10_zvstrans <- apply_scaling(iris10, attr_to_scale_info(iris10zvc_scale))