diff --git a/NAMESPACE b/NAMESPACE
index af22964..03ece9e 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -8,9 +8,4 @@ export(kmeans_estimation)
 export(kmeans_inference_1f)
 export(norm_phi_canonical_kmeans_1f)
 export(rect_hier_clusters)
-export(test_clusters_approx)
-export(test_clusters_approx_1f)
-export(test_complete_hier_clusters_approx)
-export(test_complete_hier_clusters_approx_1f)
-export(test_hier_clusters_exact)
 export(test_hier_clusters_exact_1f)
diff --git a/man/compute_S_average.Rd b/man/compute_S_average.Rd
deleted file mode 100644
index 397fcad..0000000
--- a/man/compute_S_average.Rd
+++ /dev/null
@@ -1,26 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_average}
-\alias{compute_S_average}
-\title{Computes the conditioning set for average linkage hierarchical clustering}
-\usage{
-compute_S_average(X, hcl, K, k1, k2)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set for average linkage hierarchical clustering
-}
-\keyword{internal}
diff --git a/man/compute_S_average_gencov.Rd b/man/compute_S_average_gencov.Rd
deleted file mode 100644
index 4ef6cc8..0000000
--- a/man/compute_S_average_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_average_gencov}
-\alias{compute_S_average_gencov}
-\title{Computes the conditioning set S for average linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_average_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for average linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/compute_S_centroid.Rd b/man/compute_S_centroid.Rd
deleted file mode 100644
index 0be3429..0000000
--- a/man/compute_S_centroid.Rd
+++ /dev/null
@@ -1,26 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_centroid}
-\alias{compute_S_centroid}
-\title{Computes the conditioning set for centroid linkage hierarchical clustering}
-\usage{
-compute_S_centroid(X, hcl, K, k1, k2)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set for centroid linkage hierarchical clustering
-}
-\keyword{internal}
diff --git a/man/compute_S_centroid_gencov.Rd b/man/compute_S_centroid_gencov.Rd
deleted file mode 100644
index 35b9aeb..0000000
--- a/man/compute_S_centroid_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_centroid_gencov}
-\alias{compute_S_centroid_gencov}
-\title{Computes the conditioning set S for centroid linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_centroid_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for centroid linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/compute_S_mcquitty.Rd b/man/compute_S_mcquitty.Rd
deleted file mode 100644
index dc51c92..0000000
--- a/man/compute_S_mcquitty.Rd
+++ /dev/null
@@ -1,26 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_mcquitty}
-\alias{compute_S_mcquitty}
-\title{Computes the conditioning set for McQuitty linkage hierarchical clustering (WPGMA)}
-\usage{
-compute_S_mcquitty(X, hcl, K, k1, k2)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set for McQuitty linkage hierarchical clustering (WPGMA)
-}
-\keyword{internal}
diff --git a/man/compute_S_mcquitty_gencov.Rd b/man/compute_S_mcquitty_gencov.Rd
deleted file mode 100644
index 8239aed..0000000
--- a/man/compute_S_mcquitty_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_mcquitty_gencov}
-\alias{compute_S_mcquitty_gencov}
-\title{Computes the conditioning set S for McQuitty linkage hierarchical clustering (WPGMA),
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_mcquitty_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for McQuitty linkage hierarchical clustering (WPGMA),
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/compute_S_median.Rd b/man/compute_S_median.Rd
deleted file mode 100644
index ee5b204..0000000
--- a/man/compute_S_median.Rd
+++ /dev/null
@@ -1,26 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_median}
-\alias{compute_S_median}
-\title{Computes the conditioning set for median linkage hierarchical clustering (WPGMC)}
-\usage{
-compute_S_median(X, hcl, K, k1, k2)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set for median linkage hierarchical clustering (WPGMC)
-}
-\keyword{internal}
diff --git a/man/compute_S_median_gencov.Rd b/man/compute_S_median_gencov.Rd
deleted file mode 100644
index a4fcdcb..0000000
--- a/man/compute_S_median_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_median_gencov}
-\alias{compute_S_median_gencov}
-\title{Computes the conditioning set S for median linkage hierarchical clustering (WPGMC),
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_median_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for median linkage hierarchical clustering (WPGMC),
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/compute_S_single_gencov.Rd b/man/compute_S_single_gencov.Rd
deleted file mode 100644
index 8115d7d..0000000
--- a/man/compute_S_single_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_single_gencov}
-\alias{compute_S_single_gencov}
-\title{Computes the conditioning set S for single linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_single_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for single linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/compute_S_ward.Rd b/man/compute_S_ward.Rd
deleted file mode 100644
index dc145f2..0000000
--- a/man/compute_S_ward.Rd
+++ /dev/null
@@ -1,26 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_ward}
-\alias{compute_S_ward}
-\title{Computes the conditioning set for Ward linkage hierarchical clustering}
-\usage{
-compute_S_ward(X, hcl, K, k1, k2)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set for Ward linkage hierarchical clustering
-}
-\keyword{internal}
diff --git a/man/compute_S_ward_gencov.Rd b/man/compute_S_ward_gencov.Rd
deleted file mode 100644
index eca38d6..0000000
--- a/man/compute_S_ward_gencov.Rd
+++ /dev/null
@@ -1,30 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_sets.R
-\name{compute_S_ward_gencov}
-\alias{compute_S_ward_gencov}
-\title{Computes the conditioning set S for ward linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix}
-\usage{
-compute_S_ward_gencov(X, hcl, K, k1, k2, stat)
-}
-\arguments{
-\item{X}{the n x q data set}
-
-\item{hcl}{hclust object obtained by clustering X}
-
-\item{K}{number of clusters}
-
-\item{k1}{the index of first cluster involved in the test}
-
-\item{k2}{the index of second cluster involved in the test}
-
-\item{stat}{the test statistic, \eqn{||\Sigma^{-1/2} x^T \nu||_2}}
-}
-\value{
-Returns an "Intervals" object containing the conditioning set.
-}
-\description{
-Computes the conditioning set S for ward linkage hierarchical clustering,
-w/o assuming isotropic covariance matrix
-}
-\keyword{internal}
diff --git a/man/preserve_cl.Rd b/man/preserve_cl.Rd
deleted file mode 100644
index cded3a2..0000000
--- a/man/preserve_cl.Rd
+++ /dev/null
@@ -1,22 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/util.R
-\name{preserve_cl}
-\alias{preserve_cl}
-\title{Checks if Ck, Ck' in C(x'(phi))}
-\usage{
-preserve_cl(cl, cl_phi, k1, k2)
-}
-\arguments{
-\item{cl}{clustering of x}
-
-\item{cl_phi}{clustering of x'(phi)}
-
-\item{k1, k2}{index of clusters involved in the test}
-}
-\value{
-Returns TRUE if Ck, Ck' in C(x'(phi)), and FALSE otherwise
-}
-\description{
-Checks if Ck, Ck' in C(x'(phi))
-}
-\keyword{internal}
diff --git a/man/same_cl.Rd b/man/same_cl.Rd
deleted file mode 100644
index cbc05c7..0000000
--- a/man/same_cl.Rd
+++ /dev/null
@@ -1,22 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/util.R
-\name{same_cl}
-\alias{same_cl}
-\title{Checks if two clusterings are the same up to permutation}
-\usage{
-same_cl(cl1, cl2, K)
-}
-\arguments{
-\item{cl1}{the first clustering}
-
-\item{cl2}{the second clustering}
-
-\item{K}{the number of clusters}
-}
-\value{
-Returns TRUE if they are the same, and FALSE otherwise
-}
-\description{
-Checks if two clusterings are the same up to permutation
-}
-\keyword{internal}
diff --git a/man/test_clusters_approx.Rd b/man/test_clusters_approx.Rd
deleted file mode 100644
index 0cde1b4..0000000
--- a/man/test_clusters_approx.Rd
+++ /dev/null
@@ -1,85 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_inf.R
-\name{test_clusters_approx}
-\alias{test_clusters_approx}
-\title{Monte Carlo significance test for any clustering method}
-\usage{
-test_clusters_approx(
-  X,
-  k1,
-  k2,
-  iso = TRUE,
-  sig = NULL,
-  SigInv = NULL,
-  ndraws = 2000,
-  cl_fun,
-  cl = NULL
-)
-}
-\arguments{
-\item{X}{\eqn{n} by \eqn{p} matrix containing numeric data.}
-
-\item{k1, k2}{Integers selecting the clusters to test.}
-
-\item{iso}{Boolean. If \code{TRUE}, isotropic covariance matrix model, otherwise not.}
-
-\item{sig}{Optional scalar specifying \eqn{\sigma}, relevant if \code{iso} is \code{TRUE}.}
-
-\item{SigInv}{Optional matrix specifying \eqn{\Sigma^{-1}}, relevant if \code{iso} is \code{FALSE}.}
-
-\item{ndraws}{Integer selecting the number of importance samples, default of 2000.}
-
-\item{cl_fun}{Function returning assignments to clusters 1 through \code{K}.}
-
-\item{cl}{Optionally pass in the results of calling \code{cl_fun} on your data. This is for
-efficiency and reproducibility (when the clustering function is non-deterministic).}
-}
-\value{
-\item{stat}{the test statistic: the Euclidean distance between the mean of cluster \code{k1} and the mean of cluster \code{k2}  }
-\item{pval}{the approximate p-value}
-\item{stderr}{standard error of the p-value estimate}
-\item{clusters}{the estimated cluster assignments}
-}
-\description{
-This function performs a user-specified clustering method \code{cl_fun} on the rows of a
-data matrix to obtain \code{K} clusters, and tests the null hypothesis of no difference in means
-between clusters \code{k1} and \code{k2}.
-}
-\details{
-In order to account for the fact that the clusters have been estimated from the data,
-the p-values are computed conditional on the fact that those clusters were estimated.
-This function approximates p-values via importance sampling.
-
-This function assumes that \code{cl_fun} takes a \eqn{n \times p} numeric data matrix as input
-and outputs integer assignments to clusters 1 through \code{K}.
-}
-\examples{
-# Simulates a 100 x 2 data set with three clusters
-set.seed(123)
-dat <- rbind(c(-1, 0), c(0, sqrt(3)), c(1, 0))[rep(1:3, length=100), ] +
-matrix(0.2*rnorm(200), 100, 2)
-
-# Function to run k-means clustering w/ k = 3 and 50 random starts
-km_cluster <- function(X) {
- km <- kmeans(X, 3, nstart=50)
- return(km$cluster)
-}
-
-# Cluster data using k-means
-clusters <- km_cluster(dat)
-table(rep(1:3, length=100), clusters)
-
-# tests for a difference in means between clusters 1 and 2
-# We pass in earlier k-means clustering results from earlier
-results <- test_clusters_approx(dat, k1=1, k2=2, cl_fun=km_cluster, ndraws=500, cl=clusters)
-results$stat
-results$pval
-results$stderr
-
-}
-\references{
-Lucy L. Gao et al. "Selective inference for hierarchical clustering". arXiv preprint (2020).
-}
-\seealso{
-\code{\link{test_clusters_approx_1f}} for approximate p-values for a difference in the mean of one feature.
-}
diff --git a/man/test_clusters_approx_1f.Rd b/man/test_clusters_approx_1f.Rd
deleted file mode 100644
index d1d8364..0000000
--- a/man/test_clusters_approx_1f.Rd
+++ /dev/null
@@ -1,81 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_inf.R
-\name{test_clusters_approx_1f}
-\alias{test_clusters_approx_1f}
-\title{Monte Carlo significance test (with respect to a single feature) for any clustering function}
-\usage{
-test_clusters_approx_1f(
-  X,
-  k1,
-  k2,
-  feat,
-  sig = NULL,
-  ndraws = 2000,
-  cl_fun,
-  cl = NULL
-)
-}
-\arguments{
-\item{X}{\eqn{n} by \eqn{p} matrix containing numeric data.}
-
-\item{k1, k2}{Integers selecting the clusters to test.}
-
-\item{feat}{Integer selecting the feature to test.}
-
-\item{sig}{Optional scalar specifying \eqn{\sigma}.}
-
-\item{ndraws}{Integer selecting the number of importance samples, default of 2000.}
-
-\item{cl_fun}{Function returning assignments to clusters 1 through \code{K}.}
-
-\item{cl}{Optionally pass in the results of calling \code{cl_fun} on your data. This is for
-efficiency and reproducibility (when the clustering function is non-deterministic).}
-}
-\value{
-\item{stat}{the test statistic: the Euclidean distance between the mean of cluster \code{k1} and the mean of cluster \code{k2}  }
-\item{pval}{the p-value}
-\item{clusters}{the estimated cluster assignments}
-}
-\description{
-This function performs a user-specified clustering method \code{cl_fun} on the rows of a
-data matrix to obtain \code{K} clusters, and tests the null hypothesis of no difference
-between the mean of feature \code{feat} in clusters \code{k1} and \code{k2}.
-}
-\details{
-In order to account for the fact that the clusters have been estimated from the data,
-the p-values are computed conditional on the fact that those clusters were estimated.
-This function approximates p-values via importance sampling.
-
-This function assumes that \code{cl_fun} takes a \eqn{n \times p} numeric data matrix as input
-and outputs integer assignments to clusters 1 through \code{K}.
-}
-\examples{
-# Simulates a 100 x 2 data set with three clusters
-set.seed(123)
-dat <- rbind(c(-1, 0), c(0, sqrt(3)), c(1, 0))[rep(1:3, length=100), ] +
-matrix(0.2*rnorm(200), 100, 2)
-
-# Function to run k-means clustering w/ k = 3 and 50 random starts
-km_cluster <- function(X) {
- km <- kmeans(X, 3, nstart=50)
- return(km$cluster)
-}
-
-# Cluster data using k-means
-clusters <- km_cluster(dat)
-table(rep(1:3, length=100), clusters)
-
-# tests for a difference in means between clusters 1 and 2
-results <- test_clusters_approx_1f(dat, k1=1, k2=2, feat = 1, cl_fun=km_cluster,
-ndraws=500, cl=clusters)
-results$stat
-results$pval
-results$stderr
-
-}
-\references{
-Lucy L. Gao et al. "Selective inference for hierarchical clustering". arXiv preprint (2020).
-}
-\seealso{
-\code{\link{test_clusters_approx}} for approximate p-values for a difference in the mean of any feature.
-}
diff --git a/man/test_complete_hier_clusters_approx.Rd b/man/test_complete_hier_clusters_approx.Rd
deleted file mode 100644
index 1c153f9..0000000
--- a/man/test_complete_hier_clusters_approx.Rd
+++ /dev/null
@@ -1,98 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_inf.R
-\name{test_complete_hier_clusters_approx}
-\alias{test_complete_hier_clusters_approx}
-\title{Monte Carlo significance test for complete linkage hierarchical clustering}
-\usage{
-test_complete_hier_clusters_approx(
-  X,
-  hcl,
-  K,
-  k1,
-  k2,
-  iso = TRUE,
-  sig = NULL,
-  SigInv = NULL,
-  ndraws = 2000
-)
-}
-\arguments{
-\item{X}{\eqn{n} by \eqn{p} matrix containing numeric data.}
-
-\item{hcl}{An object of the type \code{hclust} containing the hierarchical clustering of X.}
-
-\item{K}{Integer selecting the total number of clusters.}
-
-\item{k1, k2}{Integers selecting the clusters to test.}
-
-\item{iso}{Boolean. If \code{TRUE}, isotropic covariance matrix model, otherwise not.}
-
-\item{sig}{Optional scalar specifying \eqn{\sigma}, relevant if \code{iso} is \code{TRUE}.}
-
-\item{SigInv}{Optional matrix specifying \eqn{\Sigma^{-1}}, relevant if \code{iso} is \code{FALSE}.}
-
-\item{ndraws}{Integer selecting the number of importance samples, default of 2000.}
-}
-\value{
-\item{stat}{the test statistic: the Euclidean distance between the mean of cluster \code{k1} and the mean of cluster \code{k2}  }
-\item{pval}{the approximate p-value}
-\item{stderr}{estimated standard error of the p-value estimate}
-}
-\description{
-This tests the null hypothesis of no difference in means between
-clusters \code{k1} and \code{k2} at level \code{K} in a complete
-linkage hierarchical clustering. (The \code{K} clusters are numbered as per
-the results of the \code{cutree} function in the \code{stats} package.)
-}
-\details{
-Important note: Before calling \code{hclust} and this function, make sure to
-load the package \code{fastcluster}. This is because the p-value approximation
-procedure requires running hierarchical clustering on a large number of simulated
-data sets, and the version of \code{hclust} in the \code{fastcluster} package
-is much faster than the version of \code{hclust} in \code{stats}.
-
-In order to account for the fact that the clusters have been estimated from the data,
-the p-values are computed conditional on the fact that those clusters were estimated.
-This function approximates p-values via importance sampling.
-
-Currently, this function supports squared Euclidean distance as a measure of dissimilarity
-between observations. (Note that complete linkage is invariant under monotone transformations
-of the measure of dissimilarity between observations, so unsquared Euclidean distance
-would produce the same hierarchical clustering.)
-
-By default, this function assumes that the covariance matrix of the features is isotropic
-i.e. \eqn{Cov(X_i) = \sigma^2 I_p}. Setting \code{iso} to false instead assumes that
-\eqn{Cov(X_i) = \Sigma}. If known, \eqn{\sigma} can be passed in using the \code{sigma} argument
-or \eqn{\Sigma^{-1}} can be passed in the \code{SigInv} argument; otherwise, an
-estimate of \eqn{\sigma} or \eqn{\Sigma} will be used.
-}
-\examples{
-# Simulates a 100 x 2 data set with no clusters
-set.seed(1)
-dat <- matrix(rnorm(200), 100, 2)
-
-# Complete linkage hierarchical clustering
-library(fastcluster)
-hcl <- hclust(dist(dat, method="euclidean")^2, method="complete")
-
-# plot dendrograms with the 1st and 2nd clusters (cut at the third level)
-# displayed in blue and orange
-plot(hcl)
-rect_hier_clusters(hcl, k=3, which=1:2, border=c("blue", "orange"))
-
-# Monte Carlo test for a difference in means between the blue and orange clusters
-test_complete_hier_clusters_approx(X=dat, hcl=hcl, K=3, k1=1, k2=2, ndraws=1000)
-
-}
-\references{
-Lucy L. Gao et al. "Selective inference for hierarchical clustering". arXiv preprint (2020).
-}
-\seealso{
-\code{\link{rect_hier_clusters}} for visualizing clusters \code{k1} and \code{k2} in the dendrogram;
-
-\code{\link{test_hier_clusters_exact}} for exact p-values for hierarchical clustering with other linkages;
-
-\code{\link{test_clusters_approx}} for approximate p-values for a user-specified clustering function;
-
-\code{\link{test_complete_hier_clusters_approx_1f}} for approximate p-values for a difference in the mean of one feature.
-}
diff --git a/man/test_complete_hier_clusters_approx_1f.Rd b/man/test_complete_hier_clusters_approx_1f.Rd
deleted file mode 100644
index 3b94637..0000000
--- a/man/test_complete_hier_clusters_approx_1f.Rd
+++ /dev/null
@@ -1,96 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_inf.R
-\name{test_complete_hier_clusters_approx_1f}
-\alias{test_complete_hier_clusters_approx_1f}
-\title{Monte Carlo significance test (with respect to a single feature) for complete linkage hierarchical clustering}
-\usage{
-test_complete_hier_clusters_approx_1f(
-  X,
-  hcl,
-  K,
-  k1,
-  k2,
-  feat,
-  sig = NULL,
-  ndraws = 2000
-)
-}
-\arguments{
-\item{X}{\eqn{n} by \eqn{p} matrix containing numeric data.}
-
-\item{hcl}{Object of the type \code{hclust} containing the hierarchical clustering of X.}
-
-\item{K}{Integer selecting the total number of clusters.}
-
-\item{k1, k2}{Integers selecting the clusters to test.}
-
-\item{feat}{Integer selecting the feature to test.}
-
-\item{sig}{Optional scalar specifying \eqn{\sigma}.}
-
-\item{ndraws}{Integer selecting the number of importance samples, default of 2000.}
-}
-\value{
-\item{stat}{the test statistic: the absolute difference between the mean of feature \code{feat} in cluster \code{k1} and the mean of feature \code{feat} in cluster \code{k2}}
-\item{pval}{the approximate p-value}
-\item{stderr}{standard error of the p-value estimate}
-}
-\description{
-This tests the null hypothesis of no difference in means in the mean of
-feature \code{feat} between clusters \code{k1} and \code{k2} at
-level \code{K} in a complete linkage hierarchical clustering. (The \code{K}
-clusters are  numbered as per the results of the \code{cutree} function in the
-\code{stats} package.)
-}
-\details{
-Important note: Before calling \code{hclust} and this function, make sure to
-load the package \code{fastcluster}. This is because the p-value approximation
-procedure requires running hierarchical clustering on a large number of simulated
-data sets, and the version of \code{hclust} in the \code{fastcluster} package
-is much faster than the version of \code{hclust} in \code{stats}.
-
-In order to account for the fact that the clusters have been estimated from the data,
-the p-values are computed conditional on the fact that those clusters were estimated.
-This function approximates p-values via importance sampling.
-
-Currently, this function supports squared Euclidean distance as a measure of dissimilarity
-between observations. (Note that complete linkage is invariant under monotone transformations
-of the measure of dissimilarity between observations, so unsquared Euclidean distance
-would produce the same hierarchical clustering.)
-
-This function assumes that the covariance matrix of the features is isotropic
-i.e. \eqn{Cov(X_i) = \sigma^2 I_p}. If known, \eqn{\sigma} can be passed in using the \code{sigma}
-argument; otherwise, an estimate of \eqn{\sigma} will be used.
-}
-\examples{
-# Simulates a 100 x 2 data set with no clusters
-set.seed(1)
-dat <-  matrix(rnorm(200), 100, 2)
-
-# Complete linkage hierarchical clustering
-library(fastcluster)
-hcl <- hclust(dist(dat, method="euclidean")^2, method="complete")
-
-# plot dendrograms with the 1st and 2nd clusters (cut at the third split)
-# displayed in blue and orange
-plot(hcl)
-rect_hier_clusters(hcl, k=3, which=1:2, border=c("blue", "orange"))
-
-# Monte Carlo test for a difference in means between the blue and orange clusters
-# wrt the 2nd feature
-test_complete_hier_clusters_approx_1f(X=dat, hcl=hcl,
-K=3, k1=1, k2=2, feat=2, ndraws=1000)
-
-}
-\references{
-Lucy L. Gao et al. "Selective inference for hierarchical clustering". arXiv preprint (2020).
-}
-\seealso{
-\code{\link{rect_hier_clusters}} for visualizing clusters \code{k1} and \code{k2} in the dendrogram;
-
-\code{\link{test_hier_clusters_exact_1f}} for exact p-values for other linkages;
-
-\code{\link{test_clusters_approx_1f}} for approximate p-values for a user-specified clustering function;
-
-\code{\link{test_complete_hier_clusters_approx}} for approximate p-values for a difference in the mean of any feature.
-}
diff --git a/man/test_hier_clusters_exact.Rd b/man/test_hier_clusters_exact.Rd
deleted file mode 100644
index c89c0f9..0000000
--- a/man/test_hier_clusters_exact.Rd
+++ /dev/null
@@ -1,91 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/trunc_inf.R
-\name{test_hier_clusters_exact}
-\alias{test_hier_clusters_exact}
-\title{Exact significance test for hierarchical clustering}
-\usage{
-test_hier_clusters_exact(
-  X,
-  link,
-  hcl,
-  K,
-  k1,
-  k2,
-  iso = TRUE,
-  sig = NULL,
-  SigInv = NULL
-)
-}
-\arguments{
-\item{X}{\eqn{n} by \eqn{p} matrix containing numeric data.}
-
-\item{link}{String selecting the linkage. Supported options are \code{"single", "average", "centroid", "ward.D", "median"}, and \code{"mcquitty"}.}
-
-\item{hcl}{Object of the type \code{hclust} containing the hierarchical clustering of X.}
-
-\item{K}{Integer selecting the total number of clusters.}
-
-\item{k1, k2}{Integers selecting the clusters to test, as indexed by the results of \code{cutree(hcl, K)}.}
-
-\item{iso}{Boolean. If \code{TRUE}, isotropic covariance matrix model, otherwise not.}
-
-\item{sig}{Optional scalar specifying \eqn{\sigma}, relevant if \code{iso} is \code{TRUE}.}
-
-\item{SigInv}{Optional matrix specifying \eqn{\Sigma^{-1}}, relevant if \code{iso} is \code{FALSE}.}
-}
-\value{
-\item{stat}{the test statistic: the Euclidean distance between the mean of cluster \code{k1} and the mean of cluster \code{k2}  }
-\item{pval}{the p-value}
-\item{trunc}{object of the type \code{Intervals} containing the conditioning set}
-}
-\description{
-This tests the null hypothesis of no difference in means between
-clusters \code{k1} and \code{k2} at level \code{K} in a hierarchical clustering.
-(The \code{K} clusters are numbered as per the results of the \code{cutree}
-function in the \code{stats} package.)
-}
-\details{
-In order to account for the fact that the clusters have been estimated from the data,
-the p-values are computed conditional on the fact that those clusters were estimated.
-This function computes p-values exactly via an analytic characterization of the conditioning set.
-
-Currently, this function supports squared Euclidean distance as a measure of dissimilarity
-between observations, and the following six linkages: single, average, centroid, Ward,
-McQuitty (also known as WPGMA), and median (also kown as WPGMC).
-
-By default, this function assumes that the covariance matrix of the features is isotropic
-i.e. \eqn{Cov(X_i) = \sigma^2 I_p}. Setting \code{iso} to \code{FALSE} instead assumes that
-\eqn{Cov(X_i) = \Sigma}. If known, \eqn{\sigma} can be passed in using the \code{sigma} argument
-or \eqn{\Sigma^{-1}} can be passed in the \code{SigInv} argument; otherwise, an
-estimate of \eqn{\sigma} or \eqn{\Sigma} will be used.
-}
-\examples{
-# Simulates a 100 x 2 data set with three clusters
-set.seed(123)
-dat <- rbind(c(-1, 0), c(0, sqrt(3)), c(1, 0))[rep(1:3, length=100), ] +
-matrix(0.2*rnorm(200), 100, 2)
-
-# Average linkage hierarchical clustering
-hcl <- hclust(dist(dat, method="euclidean")^2, method="average")
-
-# plot dendrograms with the 1st and 2nd clusters (cut at the third split)
-# displayed in blue and orange
-plot(hcl)
-rect_hier_clusters(hcl, k=3, which=1:2, border=c("blue", "orange"))
-
-# tests for a difference in means between the blue and orange clusters
-test_hier_clusters_exact(X=dat, link="average", hcl=hcl, K=3, k1=1, k2=2)
-
-}
-\references{
-Lucy L. Gao et al. "Selective inference for hierarchical clustering". arXiv preprint (2020).
-}
-\seealso{
-\code{\link{rect_hier_clusters}} for visualizing clusters \code{k1} and \code{k2} in the dendrogram;
-
-\code{\link{test_complete_hier_clusters_approx}} for approximate p-values for complete linkage hierarchical clustering;
-
-\code{\link{test_clusters_approx}} for approximate p-values for a user-specified clustering function;
-
-\code{\link{test_hier_clusters_exact_1f}} for exact p-values for a difference in the mean of one feature.
-}