diff --git a/.github/workflows/rcmdcheck.yml b/.github/workflows/rcmdcheck.yml
index 164639528..37bb858a9 100644
--- a/.github/workflows/rcmdcheck.yml
+++ b/.github/workflows/rcmdcheck.yml
@@ -57,6 +57,11 @@ jobs:
           pak::pkg_system_requirements("rcmdcheck", execute = TRUE)
         shell: Rscript {0}
 
+      - name: Install valgrind
+        if: ${{ runner.os == 'Linux' && matrix.config.r == 'release'}}
+        run: |
+          sudo apt-get install --yes valgrind
+
       - name: Install dependencies
         run: |
           pak::local_install_dev_deps(upgrade = TRUE)
@@ -78,6 +83,11 @@ jobs:
           rcmdcheck::rcmdcheck(args = c("--no-manual", "--as-cran"), error_on = "warning", check_dir = "check")
         shell: Rscript {0}
 
+      - name: Valgrind
+        if: ${{ runner.os == 'Linux' && matrix.config.r == 'release' && always()}}
+        run: rcmdcheck::rcmdcheck(build_args = "--no-build-vignettes", args = c("--use-valgrind", "--no-codoc", "--no-manual", "--ignore-vignettes"), error_on = "error", check_dir = "check")
+        shell: Rscript {0}
+
       - name: Show testthat output
         if: always()
         run: find check -name 'testthat.Rout*' -exec cat '{}' \; || true
diff --git a/DESCRIPTION b/DESCRIPTION
index 8407f3758..28b1e824d 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: distr6
 Title: The Complete R6 Probability Distributions Interface
-Version: 1.8.0
+Version: 1.8.1
 Authors@R:
     c(person(given = "Raphael",
              family = "Sonabend",
@@ -220,6 +220,7 @@ Collate:
     'listWrappers.R'
     'makeUniqueDistributions.R'
     'measures.R'
+    'merge_cols.R'
     'mixMatrix.R'
     'mixturiseVector.R'
     'plot_continuous.R'
@@ -234,4 +235,5 @@ Collate:
     'simulateEmpiricalDistribution.R'
     'skewType.R'
     'sugar.R'
+    'transformers.R'
     'zzz.R'
diff --git a/NEWS.md b/NEWS.md
index 731aebbe8..ffdf2f633 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,7 @@
+# distr6 1.8.1
+
+* Add 'transformer' functions `pdfcdf` and `cdfpdf`, which use Rcpp to transform matrics/arrays/vectors between pdf->cdf and cdf->pdf respectively.
+
 # distr6 1.8.0
 
 * Add `Arrdist`, which generalises the `Matdist` to a three-dimensional array, useful for Bayesian predictions where the third dimension is multiple distributions.
diff --git a/R/RcppExports.R b/R/RcppExports.R
index 81026040f..e18051ed4 100644
--- a/R/RcppExports.R
+++ b/R/RcppExports.R
@@ -173,6 +173,22 @@ C_UniformKernelQuantile <- function(x, lower, logp) {
     .Call(`_distr6_C_UniformKernelQuantile`, x, lower, logp)
 }
 
+C_vec_PdfCdf <- function(x) {
+    .Call(`_distr6_C_vec_PdfCdf`, x)
+}
+
+C_vec_CdfPdf <- function(x) {
+    .Call(`_distr6_C_vec_CdfPdf`, x)
+}
+
+C_mat_PdfCdf <- function(x) {
+    .Call(`_distr6_C_mat_PdfCdf`, x)
+}
+
+C_mat_CdfPdf <- function(x) {
+    .Call(`_distr6_C_mat_CdfPdf`, x)
+}
+
 C_dpq <- function(fun, x, args, lower = TRUE, log = FALSE) {
     .Call(`_distr6_C_dpq`, fun, x, args, lower, log)
 }
diff --git a/R/SDistribution_Arrdist.R b/R/SDistribution_Arrdist.R
index bd70c1879..e6d8efd53 100644
--- a/R/SDistribution_Arrdist.R
+++ b/R/SDistribution_Arrdist.R
@@ -396,33 +396,12 @@ c.Arrdist <- function(...) {
     stop("Can't combine array distributions with different lengths on third dimension.")
   }
 
-  pdfs = .merge_arrpdf_cols(pdfs)
-  pdfs = do.call(abind::abind, list(what = pdfs, along = 1))
+  pdfs <- .merge_arrpdf_cols(pdfs)
+  pdfs <- do.call(abind::abind, list(what = pdfs, along = 1))
 
   as.Distribution(pdfs, fun = "pdf", decorators = decs)
 }
 
-.merge_arrpdf_cols <- function(pdfs) {
-  nc <- unique(viapply(pdfs, ncol))
-
-  if (length(nc) == 1) {
-    if (all(vapply(pdfs, colnames, character(nc)) == colnames(pdfs[[1]]))) {
-      return(pdfs)
-    }
-  }
-
-  cnms <- sort(unique(as.numeric(unlist(lapply(pdfs, colnames)))))
-  # new number of rows and columns
-  nc <- length(cnms)
-  nl <- dim(pdfs[[1]])[3L]
-
-  lapply(pdfs, function(.x) {
-    out <- array(0, c(nrow(.x), nc, nl), list(NULL, cnms, NULL))
-    out[, match(as.numeric(colnames(.x)), cnms), ] <- .x
-    out
-  })
-}
-
 #' @title Extract one or more Distributions from an Array distribution
 #' @description Extract a [WeightedDiscrete] or [Matdist] or [Arrdist] from a [Arrdist].
 #' @param ad [Arrdist] from which to extract Distributions.
diff --git a/R/SDistribution_Matdist.R b/R/SDistribution_Matdist.R
index 55e84d665..af2003c85 100644
--- a/R/SDistribution_Matdist.R
+++ b/R/SDistribution_Matdist.R
@@ -358,28 +358,6 @@ c.Matdist <- function(...) {
                   decorators = decs)
 }
 
-.merge_matpdf_cols <- function(pdfs) {
-
-  nc <- unique(viapply(pdfs, ncol))
-
-  if (length(nc) == 1) {
-    if (all(vapply(pdfs, colnames, character(nc)) == colnames(pdfs[[1]]))) {
-      return(pdfs)
-    }
-  }
-
-  cnms <- sort(unique(as.numeric(unlist(lapply(pdfs, colnames)))))
-  # new number of rows and columns
-  nc <- length(cnms)
-
-  lapply(pdfs, function(.x) {
-    out <- matrix(0, nrow(.x), nc, FALSE, list(NULL, cnms))
-    out[, match(as.numeric(colnames(.x)), cnms)] <- .x
-    out
-  })
-}
-
-
 #' @title Extract one or more Distributions from a Matdist
 #' @description Extract a [WeightedDiscrete] or [Matdist] from a [Matdist].
 #' @param md [Matdist] from which to extract Distributions.
diff --git a/R/getParameterSet.R b/R/getParameterSet.R
index 673bb1313..1c17cb0fb 100644
--- a/R/getParameterSet.R
+++ b/R/getParameterSet.R
@@ -685,12 +685,12 @@ getParameterSet.WeightedDiscrete <- function(object, ...) { # nolint
 
       if (length(pdfs)) {
         cdfs <- setNames(
-          lapply(pdfs, cumsum),
+          lapply(pdfs, pdfcdf),
           gsub("pdf", "cdf", names(pdfs))
         )
       } else {
         pdfs <- setNames(
-          lapply(cdfs, function(.x) c(.x[1], diff(.x))),
+          lapply(cdfs, cdfpdf),
           gsub("cdf", "pdf", names(cdfs))
         )
       }
@@ -716,9 +716,9 @@ getParameterSet.Matdist <- function(object, ...) { # nolint
       cdf <- list_element(x, "cdf")$cdf
 
       if (length(pdf)) {
-        cdf <- t(apply(pdf, 1, cumsum))
+        cdf <- pdfcdf(pdf)
       } else {
-        pdf <- t(apply(cdf, 1, function(.y) c(.y[1], diff(.y))))
+        pdf <- cdfpdf(cdf)
       }
 
       assert_cdf_matrix(cdf, pdf)
@@ -742,15 +742,9 @@ getParameterSet.Arrdist <- function(object, which.curve = 0.5, ...) { # nolint
       cdf <- list_element(x, "cdf")$cdf
 
       if (length(pdf)) {
-        cdf <- pdf
-        for (i in 2:ncol(cdf)) {
-          cdf[, i, ] <- cdf[, i, ] + cdf[, i - 1, ]
-        }
+        cdf <- pdfcdf(pdf)
       } else {
-        pdf <- cdf
-        for (i in ncol(pdf):2) {
-          pdf[, i, ] <- pdf[, i, ] - pdf[, i - 1, ]
-        }
+        pdf <- cdfpdf(cdf)
       }
 
       assert_cdf_array(cdf, pdf)
diff --git a/R/merge_cols.R b/R/merge_cols.R
new file mode 100644
index 000000000..81d93cf19
--- /dev/null
+++ b/R/merge_cols.R
@@ -0,0 +1,74 @@
+# unify column names across pdfs in
+#  matrices (merge_matpdf_cols) or arrays (merge_arrpdf_cols)
+.merge_cols <- function(arrs, fun = "pdf") {
+  if (fun == "cdf") {
+    arrs <- lapply(arrs, cdfpdf)
+  } else if (fun == "surv") {
+    arrs <- lapply(arrs, function(.x) cdfpdf(1 - .x))
+  } else if (fun != "pdf") {
+    stop(sprintf(
+      "Expected 'fun' to be 'pdf', 'cdf', or 'surv'. Got: '%s'.",
+      fun
+    ))
+  }
+
+  if (dim(arrs[[1L]]) == 2L) {
+    out <- .merge_matpdf_cols(arrs)
+  } else {
+    out <- .merge_arrpdf_cols(arrs)
+  }
+
+  if (fun == "cdf") {
+    lapply(out, pdfcdf)
+  } else if (fun == "surv") {
+    lapply(out, function(.x) 1 - pdfcdf(.x))
+  } else {
+    out
+  }
+}
+
+.merge_arrpdf_cols <- function(pdfs) {
+  if (length(unique(viapply(pdfs, function(.x) dim(.x)[[3L]]))) > 1) {
+    stop("Can only merge arrays with same length on third dimension.")
+  }
+
+  nc <- unique(viapply(pdfs, ncol))
+
+  if (length(nc) == 1) {
+    if (all(vapply(pdfs, colnames, character(nc)) == colnames(pdfs[[1]]))) {
+      return(pdfs)
+    }
+  }
+
+  cnms <- sort(unique(as.numeric(unlist(lapply(pdfs, colnames)))))
+  # new number of rows and columns
+  nc <- length(cnms)
+  nl <- dim(pdfs[[1]])[3L]
+
+  lapply(pdfs, function(.x) {
+    out <- array(0, c(nrow(.x), nc, nl), list(NULL, cnms, NULL))
+    out[, match(as.numeric(colnames(.x)), cnms), ] <- .x
+    out
+  })
+}
+
+.merge_matpdf_cols <- function(pdfs) {
+
+  nc <- unique(viapply(pdfs, ncol))
+
+  if (length(nc) == 1) {
+    if (all(vapply(pdfs, colnames, character(nc)) == colnames(pdfs[[1]]))) {
+      return(pdfs)
+    }
+  }
+
+  cnms <- sort(unique(as.numeric(unlist(lapply(pdfs, colnames)))))
+  # new number of rows and columns
+  nc <- length(cnms)
+
+  lapply(pdfs, function(.x) {
+    out <- matrix(0, nrow(.x), nc, FALSE, list(NULL, cnms))
+    out[, match(as.numeric(colnames(.x)), cnms)] <- .x
+    out
+  })
+}
\ No newline at end of file
diff --git a/R/transformers.R b/R/transformers.R
new file mode 100644
index 000000000..cf8e7b6c1
--- /dev/null
+++ b/R/transformers.R
@@ -0,0 +1,37 @@
+pdfcdf <- function(pdf) {
+  d <- dim(pdf)
+  dn <- dimnames(pdf)
+
+  if (is.null(d)) {
+    return(cumsum(pdf))
+  } else if (length(d) == 2) {
+    out <- C_mat_PdfCdf(pdf)
+  } else if (length(d) == 3) {
+    # quicker than apply with C_mat_
+    out <- aperm(apply(unname(pdf), c(1, 3), C_vec_PdfCdf), c(2, 1, 3))
+  } else {
+    stop(sprintf("Expected maximum of three dimensions but got '%s'.", length(d)))
+  }
+
+  dimnames(out) <- dn
+  out
+}
+
+cdfpdf <- function(cdf) {
+  d <- dim(cdf)
+  dn <- dimnames(cdf)
+
+  if (is.null(d)) {
+    return(c(cdf[1], diff(cdf)))
+  } else if (length(d) == 2) {
+    out <- C_mat_CdfPdf(cdf)
+  } else if (length(d) == 3) {
+    # quicker than apply with C_mat_
+    out <- aperm(apply(unname(cdf), c(1, 3), C_vec_CdfPdf), c(2, 1, 3))
+  } else {
+    stop(sprintf("Expected maximum of three dimensions but got '%s'.", length(d)))
+  }
+
+    dimnames(out) <- dn
+    out
+}
diff --git a/src/RcppExports.cpp b/src/RcppExports.cpp
index 652f6c189..380ea5b02 100644
--- a/src/RcppExports.cpp
+++ b/src/RcppExports.cpp
@@ -582,6 +582,50 @@ BEGIN_RCPP
     return rcpp_result_gen;
 END_RCPP
 }
+// C_vec_PdfCdf
+NumericVector C_vec_PdfCdf(NumericVector x);
+RcppExport SEXP _distr6_C_vec_PdfCdf(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< NumericVector >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(C_vec_PdfCdf(x));
+    return rcpp_result_gen;
+END_RCPP
+}
+// C_vec_CdfPdf
+NumericVector C_vec_CdfPdf(NumericVector x);
+RcppExport SEXP _distr6_C_vec_CdfPdf(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< NumericVector >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(C_vec_CdfPdf(x));
+    return rcpp_result_gen;
+END_RCPP
+}
+// C_mat_PdfCdf
+NumericMatrix C_mat_PdfCdf(NumericMatrix x);
+RcppExport SEXP _distr6_C_mat_PdfCdf(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< NumericMatrix >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(C_mat_PdfCdf(x));
+    return rcpp_result_gen;
+END_RCPP
+}
+// C_mat_CdfPdf
+NumericMatrix C_mat_CdfPdf(NumericMatrix x);
+RcppExport SEXP _distr6_C_mat_CdfPdf(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< NumericMatrix >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(C_mat_CdfPdf(x));
+    return rcpp_result_gen;
+END_RCPP
+}
 // C_dpq
 NumericMatrix C_dpq(std::string fun, NumericVector x, std::list<NumericVector> args, int lower, int log);
 RcppExport SEXP _distr6_C_dpq(SEXP funSEXP, SEXP xSEXP, SEXP argsSEXP, SEXP lowerSEXP, SEXP logSEXP) {
@@ -655,6 +699,10 @@ static const R_CallMethodDef CallEntries[] = {
     {"_distr6_C_UniformKernelPdf", (DL_FUNC) &_distr6_C_UniformKernelPdf, 2},
     {"_distr6_C_UniformKernelCdf", (DL_FUNC) &_distr6_C_UniformKernelCdf, 3},
     {"_distr6_C_UniformKernelQuantile", (DL_FUNC) &_distr6_C_UniformKernelQuantile, 3},
+    {"_distr6_C_vec_PdfCdf", (DL_FUNC) &_distr6_C_vec_PdfCdf, 1},
+    {"_distr6_C_vec_CdfPdf", (DL_FUNC) &_distr6_C_vec_CdfPdf, 1},
+    {"_distr6_C_mat_PdfCdf", (DL_FUNC) &_distr6_C_mat_PdfCdf, 1},
+    {"_distr6_C_mat_CdfPdf", (DL_FUNC) &_distr6_C_mat_CdfPdf, 1},
     {"_distr6_C_dpq", (DL_FUNC) &_distr6_C_dpq, 5},
     {"_distr6_C_r", (DL_FUNC) &_distr6_C_r, 3},
     {NULL, NULL, 0}
diff --git a/src/Transformers.cpp b/src/Transformers.cpp
new file mode 100644
index 000000000..9b8bd906f
--- /dev/null
+++ b/src/Transformers.cpp
@@ -0,0 +1,70 @@
+#include <Rcpp.h>
+using namespace Rcpp;
+
+// [[Rcpp::export]]
+NumericVector C_vec_PdfCdf(NumericVector x) {
+  int len = x.size();
+
+  NumericVector out(len);
+  out[0] = x[0];
+
+  for (int i = 1; i < len; i++) {
+    out[i] = x[i] + out[i - 1];
+  }
+
+  return out;
+}
+
+// [[Rcpp::export]]
+NumericVector C_vec_CdfPdf(NumericVector x) {
+  int len = x.size();
+
+  NumericVector out(len);
+  out[0] = x[0];
+
+  for (int i = (len - 1); i > 0; i--) {
+    out[i] = x[i] - x[i - 1];
+  }
+
+  return out;
+}
+
+// [[Rcpp::export]]
+NumericMatrix C_mat_PdfCdf(NumericMatrix x) {
+
+  int nc = x.ncol();
+  int nr = x.nrow();
+
+  NumericMatrix out(nr, nc);
+
+  for (int i = 0; i < nr; i++) {
+    for (int j = 0; j < nc; j++) {
+      if (j == 0) {
+        out(i, j) = x(i, j);
+      } else {
+        out(i, j) = x(i, j) + out(i, j - 1);
+      }
+    }
+  }
+  return out;
+}
+
+// [[Rcpp::export]]
+NumericMatrix C_mat_CdfPdf(NumericMatrix x) {
+
+  int nr = x.nrow();
+  int nc = x.ncol();
+
+  NumericMatrix out(nr, nc);
+
+  for (int i = 0; i < nr; i++) {
+    for (int j = (nc - 1); j >= 0; j--) {
+      if (j == 0) {
+        out(i, j) = x(i, j);
+      } else {
+        out(i, j) = x(i, j) - x(i, j - 1);
+      }
+    }
+  }
+  return out;
+}
diff --git a/tests/testthat/test-dparse.R b/tests/testthat/test-dparse.R
index 908d2c30e..5196d0fe4 100644
--- a/tests/testthat/test-dparse.R
+++ b/tests/testthat/test-dparse.R
@@ -38,7 +38,7 @@ test_that("ShortName, ClassName and Alias are unique between distributions", {
   expect_equal(length(d6), length(unique(d6)))
 })
 
-test_that("Every distribution is created with it's propper R6 class", {
+test_that("Every distribution is created with it's proper R6 class", {
   # Get calls
   d6 <- listDistributions()[,(ids = paste(tolower(ShortName), tolower(ClassName), tolower(Alias), sep = ", "))]
   calls <- strsplit(d6, ",")
diff --git a/tests/testthat/test-sdistribution-Arrdist.R b/tests/testthat/test-sdistribution-Arrdist.R
index d1bf4e9ba..096dff9a7 100644
--- a/tests/testthat/test-sdistribution-Arrdist.R
+++ b/tests/testthat/test-sdistribution-Arrdist.R
@@ -67,6 +67,12 @@ test_that("c.Arrdist", {
   expect_equal(dim(r), c(50, 60))
   expect_true(all(r <= 20))
   expect_true(all(r >= 1))
+
+  arr4pdf = array(runif(200), c(20, 10, 1))
+  arr4pdf = arr4pdf / sum(arr4pdf)
+  colnames(arr4pdf) = 1:10
+  arr4 = as.Distribution(arr4pdf, fun = "pdf")
+  expect_error(c(arr1, arr4), "Can't combine")
 })
 
 test_that("Arrdist basics", {
diff --git a/tests/testthat/test-transformers.R b/tests/testthat/test-transformers.R
new file mode 100644
index 000000000..2e98d8efa
--- /dev/null
+++ b/tests/testthat/test-transformers.R
@@ -0,0 +1,30 @@
+test_that("vector transformers", {
+  pdf <- c(0.1, 0.2, 0.7)
+  cdf <- cumsum(pdf)
+
+  expect_equal(pdfcdf(pdf), cdf)
+  expect_equal(cdfpdf(cdf), pdf)
+})
+
+test_that("matrix transformers", {
+  pdf <- matrix(c(0.1, 0.2, 0.7, 0.2, 0.2, 0.6), 2, 3, TRUE)
+  cdf <- t(apply(pdf, 1, cumsum))
+
+  expect_equal(pdfcdf(pdf), cdf)
+  expect_equal(cdfpdf(cdf), pdf)
+})
+
+test_that("array transformers", {
+  pdf <- matrix(c(0.1, 0.2, 0.7, 0.2, 0.2, 0.6), 2, 3, TRUE)
+  cdf <- t(apply(pdf, 1, cumsum))
+  pdf <- array(pdf, c(2, 3, 5))
+  cdf <- array(cdf, c(2, 3, 5))
+
+  expect_equal(pdfcdf(pdf), cdf)
+  expect_equal(cdfpdf(cdf), pdf)
+})
+
+test_that("transformers error when expected", {
+  expect_error(pdfcdf(array(dim = c(1, 1, 1, 1))), "Expected maximum")
+  expect_error(cdfpdf(array(dim = c(1, 1, 1, 1))), "Expected maximum")
+})