tests/testthat/test-evalCand.R
In fionarhuang/treeclimbR: An algorithm to find optimal signal levels in a tree
test_that(".pseudoLeaf works", {
library(TreeSummarizedExperiment)
library(ggtree)
data(tinyTree)
set.seed(1)
pv <- runif(19, 0, 1)
pv[c(seq_len(5), 13, 14, 18)] <- runif(8, 0, 0.001)
fc <- sample(c(-1, 1), 19, replace = TRUE)
fc[c(seq_len(3), 13, 14)] <- 1
fc[c(4, 5, 18)] <- -1
df <- data.frame(node = seq_len(19),
pvalue = pv,
foldChange = fc)
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df,
node_column = "node", p_column = "pvalue"),
seq(1, 10))
## 2,3 are NA - replace with next level
df2 <- df
df2$pvalue[df2$node %in% c(2, 3)] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(1, 14, seq(4, 10)))
## only 2 is NA - gets removed
df2 <- df
df2$pvalue[df2$node %in% 2] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(1, 3, seq(4, 10)))
## 4,5,6 are NA - replace 4 and 5 with 18 and remove 6
df2 <- df
df2$pvalue[df2$node %in% c(4, 5, 6)] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(seq(1, 3), 18, seq(7, 10)))
})
test_that("evalCand works", {
## Generate some data
library(TreeSummarizedExperiment)
data(tinyTree)
set.seed(1)
pv <- runif(19, 0, 1)
pv[c(seq_len(5), 13, 14, 18)] <- runif(8, 0, 0.001)
fc <- sample(c(-1, 1), 19, replace = TRUE)
fc[c(seq_len(3), 13, 14)] <- 1
fc[c(4, 5, 18)] <- -1
df <- data.frame(node = seq_len(19), pvalue = pv,
foldChange = fc, feature = "gene1")
ll <- getCand(tree = tinyTree, score_data = df,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange")
## One with NAs
df2 <- df
df2$feature <- "gene2"
df2$pvalue[df2$node %in% c(4, 5, 6)] <- NA
llna <- getCand(tree = tinyTree, score_data = df2,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 1e-4,
pct_na = 0.25, message = FALSE)
## Stricter threshold on pct_na -> however, node 18 will still be chosen
## since it does not have any descendants with non-NA p-value and thus
## is considered a (pseudo)-leaf
llnastrict <- getCand(tree = tinyTree, score_data = df2,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 0.05,
pct_na = 0.75, message = FALSE)
## Stricter threshold on pct_na, and now leaf 4 has a valid p-value, so
## node 18 is not a leaf, but will not be chosen
df2p <- df
df2p$pvalue[df2p$node %in% c(5, 6)] <- NA
llnastrict2 <- getCand(tree = tinyTree, score_data = df2p,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 0.05,
pct_na = 0.75, message = FALSE)
## One where there are different candidates generated
df3 <- df
df3$foldChange[df3$node %in% c(17, 7, 19)] <- -1
lldiff <- getCand(tree = tinyTree, score_data = df3,
t = c(0.01, 0.05, 0.1, 0.25, 0.95),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 1,
pct_na = 0.5, message = FALSE)
## Check that the function returns error for invalid input
## -------------------------------------------------------------------------
.args <- list(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
args <- .args
args$tree <- 1
expect_error(do.call(evalCand, args),
"'tree' must be of class 'phylo'")
args <- .args
args$type <- 1
expect_error(do.call(evalCand, args),
"'arg' must be NULL or a character vector")
args$type <- "missing"
expect_error(do.call(evalCand, args),
"'arg' should be one of")
args <- .args
args$levels <- c(1, 2)
expect_error(do.call(evalCand, args),
"'levels' must be of class 'list'")
args <- .args
args$score_data <- c(1, 2)
expect_error(do.call(evalCand, args),
"'score_data' must be of class 'data.frame'")
args$type <- "multiple"
expect_error(do.call(evalCand, args),
"'score_data' must be of class 'list'")
args <- .args
args$node_column <- 1
expect_error(do.call(evalCand, args),
"'node_column' must be of class 'character'")
args$node_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'node_column' must have length 1")
args$node_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$p_column <- 1
expect_error(do.call(evalCand, args),
"'p_column' must be of class 'character'")
args$p_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'p_column' must have length 1")
args$p_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$sign_column <- 1
expect_error(do.call(evalCand, args),
"'sign_column' must be of class 'character'")
args$sign_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'sign_column' must have length 1")
args$sign_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$type <- "multiple"
args$score_data <- list(args$score_data)
args$feature_column <- 1
expect_error(do.call(evalCand, args),
"'feature_column' must be of class 'character'")
args$feature_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'feature_column' must have length 1")
args$feature_column <- "missing"
expect_error(do.call(evalCand, args),
"feature_column %in%",
fixed = TRUE)
args <- .args
args$method <- 1
expect_error(do.call(evalCand, args),
"'method' must be of class 'character'")
args$method <- c("BH", "holm")
expect_error(do.call(evalCand, args),
"'method' must have length 1")
args <- .args
args$limit_rej <- "x"
expect_error(do.call(evalCand, args),
"'limit_rej' must be of class 'numeric'")
args$limit_rej <- c(0.1, 0.2)
expect_error(do.call(evalCand, args),
"'limit_rej' must have length 1")
args$limit_rej <- 2
expect_error(do.call(evalCand, args),
"'limit_rej' must be within [0,1]", fixed = TRUE)
args <- .args
args$use_pseudo_leaf <- "x"
expect_error(do.call(evalCand, args),
"'use_pseudo_leaf' must be of class 'logical'")
args$use_pseudo_leaf <- c(TRUE, FALSE)
expect_error(do.call(evalCand, args),
"'use_pseudo_leaf' must have length 1")
args <- .args
args$message <- "x"
expect_error(do.call(evalCand, args),
"'message' must be of class 'logical'")
args$message <- c(TRUE, FALSE)
expect_error(do.call(evalCand, args),
"'message' must have length 1")
## Check that the function works as expected for valid input
## -------------------------------------------------------------------------
## Single
out <- evalCand(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.15, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 5, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
ll$score_data[match(out$output$node, ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(sum(out$output$adj.p <= 0.05), 2)
## Single - different candidates
out <- evalCand(tree = tinyTree, type = "single",
levels = lldiff$candidate_list,
score_data = lldiff$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.95, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.95)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.10, upper_t = 1, is_valid = TRUE,
method = "BH", limit_rej = 0.95, level_name = "0.1",
best = FALSE, rej_leaf = 10, rej_node = 7,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(out$level_info[5, ],
data.frame(t = 0.95, upper_t = 1, is_valid = TRUE,
method = "BH", limit_rej = 0.95, level_name = "0.95",
best = TRUE, rej_leaf = 10, rej_node = 4,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(lldiff$candidate_list$`0.95`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
lldiff$score_data[match(out$output$node, lldiff$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
## Single - NA values
out <- evalCand(tree = tinyTree, type = "single",
levels = llna$candidate_list,
score_data = llna$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.1, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 6, rej_node = 5,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llna$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llna$score_data[match(out$output$node, llna$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$candidate_best[[1]], c(1, 2, 3, 18, 7, 8, 9, 10))
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$output$node[out$output$signal.node], c(1, 2, 3, 18, 10))
## Single - NA values with more strict requirements on pct_na
out <- evalCand(tree = tinyTree, type = "single",
levels = llnastrict$candidate_list,
score_data = llnastrict$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.3, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 5, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llnastrict$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llnastrict$score_data[match(out$output$node, llnastrict$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$candidate_best[[1]], c(18, 7, 8, 9, 10, 13))
expect_equal(out$output$node[out$output$signal.node], c(18, 13))
## Single - NA values with more strict requirements on pct_na
out <- evalCand(tree = tinyTree, type = "single",
levels = llnastrict2$candidate_list,
score_data = llnastrict2$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.2, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 4, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llnastrict2$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llnastrict2$score_data[match(out$output$node, llnastrict2$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$candidate_best[[1]], c(4, 7, 8, 9, 10, 13))
expect_equal(out$output$node[out$output$signal.node], c(4, 13))
## Single - too low FDR
expect_error({
out <- evalCand(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 1e-10, use_pseudo_leaf = FALSE,
message = FALSE)
}, "No valid level could be found")
## Multiple
out <- evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = "feature"))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.1, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 10, rej_node = 6,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`) +
length(llna$candidate_list$`0.01`))
expect_equal(out$output[out$output$feature == "gene1", seq_len(7)],
ll$score_data[match(out$output$node[out$output$feature == "gene1"],
ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
## Multiple - use pseudo leaf
out <- evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = TRUE,
message = TRUE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = "feature"))
expect_equal(out$level_info$rej_pseudo_leaf, rep(9, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.08, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 9, rej_node = 6,
rej_pseudo_leaf = 9, rej_pseudo_node = 5),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`) +
length(llna$candidate_list$`0.01`))
expect_equal(out$output[out$output$feature == "gene1", seq_len(7)],
ll$score_data[match(out$output$node[out$output$feature == "gene1"],
ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
## Multiple - different levels for different features
ll1 <- ll
ll1$candidate_list <- ll1$candidate_list[seq_len(3)]
expect_error(evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll1$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE),
"The names of elements in 'levels' are different")
## Multiple - no feature column
expect_warning(evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE),
"To distinguish results from different features")
})
fionarhuang/treeclimbR documentation built on Oct. 11, 2024, 6:19 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
test_that(".pseudoLeaf works", {
library(TreeSummarizedExperiment)
library(ggtree)
data(tinyTree)
set.seed(1)
pv <- runif(19, 0, 1)
pv[c(seq_len(5), 13, 14, 18)] <- runif(8, 0, 0.001)
fc <- sample(c(-1, 1), 19, replace = TRUE)
fc[c(seq_len(3), 13, 14)] <- 1
fc[c(4, 5, 18)] <- -1
df <- data.frame(node = seq_len(19),
pvalue = pv,
foldChange = fc)
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df,
node_column = "node", p_column = "pvalue"),
seq(1, 10))
## 2,3 are NA - replace with next level
df2 <- df
df2$pvalue[df2$node %in% c(2, 3)] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(1, 14, seq(4, 10)))
## only 2 is NA - gets removed
df2 <- df
df2$pvalue[df2$node %in% 2] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(1, 3, seq(4, 10)))
## 4,5,6 are NA - replace 4 and 5 with 18 and remove 6
df2 <- df
df2$pvalue[df2$node %in% c(4, 5, 6)] <- NA
expect_equal(.pseudoLeaf(tree = tinyTree, score_data = df2,
node_column = "node", p_column = "pvalue"),
c(seq(1, 3), 18, seq(7, 10)))
})
test_that("evalCand works", {
## Generate some data
library(TreeSummarizedExperiment)
data(tinyTree)
set.seed(1)
pv <- runif(19, 0, 1)
pv[c(seq_len(5), 13, 14, 18)] <- runif(8, 0, 0.001)
fc <- sample(c(-1, 1), 19, replace = TRUE)
fc[c(seq_len(3), 13, 14)] <- 1
fc[c(4, 5, 18)] <- -1
df <- data.frame(node = seq_len(19), pvalue = pv,
foldChange = fc, feature = "gene1")
ll <- getCand(tree = tinyTree, score_data = df,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange")
## One with NAs
df2 <- df
df2$feature <- "gene2"
df2$pvalue[df2$node %in% c(4, 5, 6)] <- NA
llna <- getCand(tree = tinyTree, score_data = df2,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 1e-4,
pct_na = 0.25, message = FALSE)
## Stricter threshold on pct_na -> however, node 18 will still be chosen
## since it does not have any descendants with non-NA p-value and thus
## is considered a (pseudo)-leaf
llnastrict <- getCand(tree = tinyTree, score_data = df2,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 0.05,
pct_na = 0.75, message = FALSE)
## Stricter threshold on pct_na, and now leaf 4 has a valid p-value, so
## node 18 is not a leaf, but will not be chosen
df2p <- df
df2p$pvalue[df2p$node %in% c(5, 6)] <- NA
llnastrict2 <- getCand(tree = tinyTree, score_data = df2p,
t = c(seq(0.01, 0.05, by = 0.01), seq(0.1, 1, by = 0.05)),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 0.05,
pct_na = 0.75, message = FALSE)
## One where there are different candidates generated
df3 <- df
df3$foldChange[df3$node %in% c(17, 7, 19)] <- -1
lldiff <- getCand(tree = tinyTree, score_data = df3,
t = c(0.01, 0.05, 0.1, 0.25, 0.95),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange", threshold = 1,
pct_na = 0.5, message = FALSE)
## Check that the function returns error for invalid input
## -------------------------------------------------------------------------
.args <- list(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
args <- .args
args$tree <- 1
expect_error(do.call(evalCand, args),
"'tree' must be of class 'phylo'")
args <- .args
args$type <- 1
expect_error(do.call(evalCand, args),
"'arg' must be NULL or a character vector")
args$type <- "missing"
expect_error(do.call(evalCand, args),
"'arg' should be one of")
args <- .args
args$levels <- c(1, 2)
expect_error(do.call(evalCand, args),
"'levels' must be of class 'list'")
args <- .args
args$score_data <- c(1, 2)
expect_error(do.call(evalCand, args),
"'score_data' must be of class 'data.frame'")
args$type <- "multiple"
expect_error(do.call(evalCand, args),
"'score_data' must be of class 'list'")
args <- .args
args$node_column <- 1
expect_error(do.call(evalCand, args),
"'node_column' must be of class 'character'")
args$node_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'node_column' must have length 1")
args$node_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$p_column <- 1
expect_error(do.call(evalCand, args),
"'p_column' must be of class 'character'")
args$p_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'p_column' must have length 1")
args$p_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$sign_column <- 1
expect_error(do.call(evalCand, args),
"'sign_column' must be of class 'character'")
args$sign_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'sign_column' must have length 1")
args$sign_column <- "missing"
expect_error(do.call(evalCand, args),
"all(c(node_column, p_column, sign_column) %in%",
fixed = TRUE)
args <- .args
args$type <- "multiple"
args$score_data <- list(args$score_data)
args$feature_column <- 1
expect_error(do.call(evalCand, args),
"'feature_column' must be of class 'character'")
args$feature_column <- c("node", "pvalue")
expect_error(do.call(evalCand, args),
"'feature_column' must have length 1")
args$feature_column <- "missing"
expect_error(do.call(evalCand, args),
"feature_column %in%",
fixed = TRUE)
args <- .args
args$method <- 1
expect_error(do.call(evalCand, args),
"'method' must be of class 'character'")
args$method <- c("BH", "holm")
expect_error(do.call(evalCand, args),
"'method' must have length 1")
args <- .args
args$limit_rej <- "x"
expect_error(do.call(evalCand, args),
"'limit_rej' must be of class 'numeric'")
args$limit_rej <- c(0.1, 0.2)
expect_error(do.call(evalCand, args),
"'limit_rej' must have length 1")
args$limit_rej <- 2
expect_error(do.call(evalCand, args),
"'limit_rej' must be within [0,1]", fixed = TRUE)
args <- .args
args$use_pseudo_leaf <- "x"
expect_error(do.call(evalCand, args),
"'use_pseudo_leaf' must be of class 'logical'")
args$use_pseudo_leaf <- c(TRUE, FALSE)
expect_error(do.call(evalCand, args),
"'use_pseudo_leaf' must have length 1")
args <- .args
args$message <- "x"
expect_error(do.call(evalCand, args),
"'message' must be of class 'logical'")
args$message <- c(TRUE, FALSE)
expect_error(do.call(evalCand, args),
"'message' must have length 1")
## Check that the function works as expected for valid input
## -------------------------------------------------------------------------
## Single
out <- evalCand(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.15, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 5, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
ll$score_data[match(out$output$node, ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(sum(out$output$adj.p <= 0.05), 2)
## Single - different candidates
out <- evalCand(tree = tinyTree, type = "single",
levels = lldiff$candidate_list,
score_data = lldiff$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.95, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.95)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.10, upper_t = 1, is_valid = TRUE,
method = "BH", limit_rej = 0.95, level_name = "0.1",
best = FALSE, rej_leaf = 10, rej_node = 7,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(out$level_info[5, ],
data.frame(t = 0.95, upper_t = 1, is_valid = TRUE,
method = "BH", limit_rej = 0.95, level_name = "0.95",
best = TRUE, rej_leaf = 10, rej_node = 4,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(lldiff$candidate_list$`0.95`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
lldiff$score_data[match(out$output$node, lldiff$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
## Single - NA values
out <- evalCand(tree = tinyTree, type = "single",
levels = llna$candidate_list,
score_data = llna$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.1, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 6, rej_node = 5,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llna$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llna$score_data[match(out$output$node, llna$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$candidate_best[[1]], c(1, 2, 3, 18, 7, 8, 9, 10))
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$output$node[out$output$signal.node], c(1, 2, 3, 18, 10))
## Single - NA values with more strict requirements on pct_na
out <- evalCand(tree = tinyTree, type = "single",
levels = llnastrict$candidate_list,
score_data = llnastrict$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.3, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 5, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llnastrict$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llnastrict$score_data[match(out$output$node, llnastrict$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$candidate_best[[1]], c(18, 7, 8, 9, 10, 13))
expect_equal(out$output$node[out$output$signal.node], c(18, 13))
## Single - NA values with more strict requirements on pct_na
out <- evalCand(tree = tinyTree, type = "single",
levels = llnastrict2$candidate_list,
score_data = llnastrict2$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.1, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.1)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = NULL))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[3, ],
data.frame(t = 0.03, upper_t = 0.2, is_valid = TRUE,
method = "BH", limit_rej = 0.1, level_name = "0.03",
best = TRUE, rej_leaf = 4, rej_node = 2,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(llnastrict2$candidate_list$`0.03`))
expect_equal(as.data.frame(out$output)[, seq_len(7)],
llnastrict2$score_data[match(out$output$node, llnastrict2$score_data$node), seq_len(7)],
ignore_attr = TRUE)
expect_equal(out$output$adj.p, p.adjust(out$output$pvalue, method = "BH"))
expect_equal(out$candidate_best[[1]], c(4, 7, 8, 9, 10, 13))
expect_equal(out$output$node[out$output$signal.node], c(4, 13))
## Single - too low FDR
expect_error({
out <- evalCand(tree = tinyTree, type = "single",
levels = ll$candidate_list, score_data = ll$score_data,
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 1e-10, use_pseudo_leaf = FALSE,
message = FALSE)
}, "No valid level could be found")
## Multiple
out <- evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = "feature"))
expect_equal(out$level_info$rej_pseudo_leaf, rep(NA, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.1, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 10, rej_node = 6,
rej_pseudo_leaf = NA, rej_pseudo_node = NA),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`) +
length(llna$candidate_list$`0.01`))
expect_equal(out$output[out$output$feature == "gene1", seq_len(7)],
ll$score_data[match(out$output$node[out$output$feature == "gene1"],
ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
## Multiple - use pseudo leaf
out <- evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = TRUE,
message = TRUE)
expect_type(out, "list")
expect_named(out, c("candidate_best", "output", "candidate_list",
"level_info", "FDR", "method", "column_info"))
expect_type(out$candidate_best, "list")
expect_s3_class(out$output, "data.frame")
expect_type(out$candidate_list, "list")
expect_s3_class(out$level_info, "data.frame")
expect_equal(out$FDR, 0.05)
expect_equal(out$method, "BH")
expect_equal(out$column_info,
list(node_column = "node", p_column = "pvalue",
sign_column = "foldChange", feature_column = "feature"))
expect_equal(out$level_info$rej_pseudo_leaf, rep(9, nrow(out$level_info)))
expect_equal(out$level_info[1, ],
data.frame(t = 0.01, upper_t = 0.08, is_valid = TRUE,
method = "BH", limit_rej = 0.05, level_name = "0.01",
best = TRUE, rej_leaf = 9, rej_node = 6,
rej_pseudo_leaf = 9, rej_pseudo_node = 5),
ignore_attr = TRUE)
expect_equal(nrow(out$output), length(ll$candidate_list$`0.01`) +
length(llna$candidate_list$`0.01`))
expect_equal(out$output[out$output$feature == "gene1", seq_len(7)],
ll$score_data[match(out$output$node[out$output$feature == "gene1"],
ll$score_data$node), seq_len(7)],
ignore_attr = TRUE)
## Multiple - different levels for different features
ll1 <- ll
ll1$candidate_list <- ll1$candidate_list[seq_len(3)]
expect_error(evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll1$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = "feature", method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE),
"The names of elements in 'levels' are different")
## Multiple - no feature column
expect_warning(evalCand(tree = tinyTree, type = "multiple",
levels = list(gene1 = ll$candidate_list,
gene2 = llna$candidate_list),
score_data = list(gene1 = ll$score_data,
gene2 = llna$score_data),
node_column = "node", p_column = "pvalue",
sign_column = "foldChange",
feature_column = NULL, method = "BH",
limit_rej = 0.05, use_pseudo_leaf = FALSE,
message = FALSE),
"To distinguish results from different features")
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.