R/plot_taxa.R
In EDIorg/ecocomDP: Tools to Create, Use, and Convert ecocomDP Data
Defines functions
plot_sites
plot_taxa_abund
plot_taxa_occur_freq
plot_taxa_rank
plot_taxa_shared_sites
plot_taxa_sample_time
plot_sample_space_time
plot_taxa_diversity
plot_taxa_accum_time
plot_taxa_accum_sites
format_for_comm_plots
Documented in
plot_sample_space_time
plot_sites
plot_taxa_abund
plot_taxa_accum_sites
plot_taxa_accum_time
plot_taxa_diversity
plot_taxa_occur_freq
plot_taxa_rank
plot_taxa_sample_time
plot_taxa_shared_sites
# Format dataset for community plotting functions
#
# @param observation (tbl_df, tbl, data.frame) The observation table of ecocomDP
# @param id (character) Identifier of dataset to be used in plot subtitles.
#
# @details Downstream plotting functions are based on \href{https://github.com/sokole/ltermetacommunities/tree/master/Group2-explore-data}{LTER Metacommunities code} and use their intermediate data input format.
#
# @return (tbl_df, tbl, data.frame) Tabular data of \code{id} in a format compatible with plotting functions
#
format_for_comm_plots <- function(observation, id) {
id <- id
# Constraints
varname <- unique(observation[c("variable_name", "unit")]) # Can only handle one variable
if (nrow(varname) > 1) {
warning("Input 'observation' has > 1 unique variable_name and unit ",
"combination. Only the first will be used (",
varname$variable_name, ", ", varname$unit, ").", call. = FALSE)
if (is.na(varname$unit[1])) {
observation <- dplyr::filter(observation,
variable_name == varname$variable_name[1] &
is.na(unit))
} else {
observation <- dplyr::filter(observation,
variable_name == varname$variable_name[1] &
unit == varname$unit[1])
}
}
# Methods used by this processing don't require aggregation of non-unique observations, so drop duplicates without warning
dups <- observation %>% dplyr::select(-observation_id, -value, -event_id) %>% duplicated()
if (any(dups)) {
observation <- observation[!dups, ]
}
# Use dates
if (is.character(observation$datetime)) {
warning("Input datetimes are character strings. Accuracy may be improved",
" if inputs are parsed before calling plot_data().", call. = FALSE)
}
observation$datetime <- lubridate::date(observation$datetime)
obs <- observation
taxon_count <- data.frame(
OBSERVATION_TYPE = "TAXON_COUNT",
SITE_ID = obs$location_id,
DATE = obs$datetime,
VARIABLE_NAME = obs$taxon_id,
VARIABLE_UNITS = obs$variable_name,
VALUE = obs$value,
stringsAsFactors = FALSE)
dslong <- taxon_count %>% # long form
dplyr::mutate_at(dplyr::vars(c(VALUE)), as.numeric) %>%
dplyr::mutate_at(dplyr::vars(SITE_ID), as.character)
dswide <- dslong %>% # wide form
dplyr::select(-VARIABLE_UNITS) %>%
tidyr::pivot_wider(
names_from = VARIABLE_NAME,
values_from = VALUE,
values_fn = mean) #average across non-unique observations.
res <- list(
id = id,
dslong = dslong,
dswide = dswide)
return(res)
}
#' Plot taxa accumulation by site accumulation
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_accum_sites(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_accum_sites()
#'
#' # Plot from the observation table directly
#' plot_taxa_accum_sites(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_accum_sites(ants_L1)
#'
plot_taxa_accum_sites <- function(data,
id = NA_character_,
alpha = 1,
observation = NULL) {
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
# Calculate cumulative number of taxa
calc_cuml_taxa_space <- function(ex) {
taxa.s.list <- list()
sites <- unique(ex$SITE_ID)
for(t in 1:length(unique(ex$SITE_ID))){ # unique taxa found in each year
tmp.dat <- subset(ex, ex$SITE_ID == sites[t])
tmp.dat.pres <- subset(tmp.dat, tmp.dat$VALUE > 0)
taxa.s.list[[t]] <- unique(tmp.dat.pres$VARIABLE_NAME)
}
cuml.taxa.space <- list() # cumulative list of taxa over space
cuml.taxa.space[[1]] <- taxa.s.list[[1]]
if (length(unique(ex$SITE_ID)) > 1) {
for(t in 2:length(unique(ex$SITE_ID))){ # list cumulative taxa, with duplicates
cuml.taxa.space[[t]] <- c(cuml.taxa.space[[t - 1]], taxa.s.list[[t]])
}
}
cuml.taxa.space <- lapply(cuml.taxa.space, function(x) {unique(x)}) # rm duplicates
cuml.no.taxa.space <- data.frame("site" = unique(ex$SITE_ID)) # total unique taxa over space
cuml.no.taxa.space$no.taxa <- unlist(lapply(cuml.taxa.space, function(x) {length(x)}))
return(cuml.no.taxa.space)
}
comm.dat <- ds$dslong %>% arrange(SITE_ID) # order by site
no.taxa.space <- calc_cuml_taxa_space(comm.dat)
no.taxa.space$no.site <- as.numeric(rownames(no.taxa.space))
# Plot
p <- ggplot2::ggplot(data = no.taxa.space, ggplot2::aes(x = no.site, y = no.taxa)) +
ggplot2::geom_point() +
ggplot2::geom_line() +
ggplot2::labs(title = "Taxa accumulation by site accumulation", subtitle = ds$id) +
ggplot2::xlab("Cumulative number of sites") +
ggplot2::ylab(paste0("Cumulative number of taxa")) +
ggplot2::theme_bw()
return(p)
}
#' Plot taxa accumulation through time
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_accum_time(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_accum_time()
#'
#' # Plot from the observation table directly
#' plot_taxa_accum_time(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_accum_time(ants_L1)
#'
plot_taxa_accum_time <- function(data,
id = NA_character_,
alpha = 1,
observation = NULL){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
# Calculate cumulative number of taxa
cuml.taxa.fun <- function(ex){
taxa.t.list <- list()
dates <- unique(ex$DATE)
for(t in 1:length(unique(ex$DATE))) { # unique taxa found in each year
tmp.dat <- subset(ex, ex$DATE == dates[t])
tmp.dat.pres <- subset(tmp.dat, tmp.dat$VALUE > 0)
taxa.t.list[[t]] <- unique(tmp.dat.pres$VARIABLE_NAME)
}
cuml.taxa <- list() # cumulative taxa through time
cuml.taxa[[1]] <- taxa.t.list[[1]]
if (length(unique(ex$DATE)) > 1) { # create list of the cumulative taxa, with duplicates
for(t in 2:length(unique(ex$DATE))){
cuml.taxa[[t]] <- c(cuml.taxa[[t - 1]], taxa.t.list[[t]])
}
}
cuml.taxa <- lapply(cuml.taxa, function(x){unique(x)}) # rm duplicates
cuml.no.taxa <- data.frame("year" = unique(ex$DATE), stringsAsFactors = FALSE) # number of unique taxa through time
cuml.no.taxa$no.taxa <- unlist(lapply(cuml.taxa, function(x){length(x)}))
return(cuml.no.taxa)
}
cuml.taxa.all.sites <- cuml.taxa.fun(
ex = ds$dslong %>% dplyr::arrange(DATE)) # taxa accumulation across all sites pooled together
# Plot
p <- ggplot2::ggplot() +
ggplot2::geom_point(
data = cuml.taxa.all.sites,
ggplot2::aes(x = year, y = no.taxa, fill = "")) +
ggplot2::geom_line(
data = cuml.taxa.all.sites,
ggplot2::aes(x = year, y = no.taxa)) +
ggplot2::labs(title = "Accumulation of taxa through time", subtitle = ds$id) +
ggplot2::xlab("Year") +
ggplot2::ylab(paste0("Cumulative number of taxa")) +
ggplot2::scale_x_date(date_labels = "%Y", date_breaks = "1 year", date_minor_breaks = "1 month") +
ggplot2::guides(fill = "none") +
ggplot2::ylim(c(0, max(cuml.taxa.all.sites$no.taxa))) +
ggplot2::theme_bw() +
theme(axis.text.x = element_text(angle = 45, hjust=1))
return(p)
}
#' Plot diversity (taxa richness) through time
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param time_window_size (character) Define the time window over which to aggregate observations for calculating richness. Can be: "day" or "year"
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_diversity(dataset)
#'
#' # Plot the dataset with observations aggregated by year
#' plot_taxa_diversity(dataset, time_window_size = "year")
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(
#' lubridate::as_date(datetime) > "2007-01-01") %>%
#' plot_taxa_diversity()
#'
#' # Plot from the observation table directly
#' plot_taxa_diversity(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_diversity(ants_L1)
#'
plot_taxa_diversity <- function(data,
id = NA_character_,
time_window_size = "day",
observation = NULL,
alpha = 1){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# Richness by location id
richness_by_location <- observation %>%
dplyr::filter(value > 0) %>%
dplyr::select(
.data$location_id,
.data$datetime,
.data$taxon_id) %>%
dplyr::distinct() %>%
dplyr::mutate(
datetime = dplyr::case_when(
grepl("^(?i)d", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
as.character(),
grepl("^(?i)m", time_window_size) ~ paste0(
.data$datetime %>% lubridate::as_date() %>% lubridate::year(),"-",
.data$datetime %>% lubridate::as_date() %>% lubridate::month(),"-01"),
grepl("^(?i)y", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
lubridate::year() %>%
as.character())) %>%
dplyr::group_by(.data$location_id, .data$datetime) %>%
dplyr::summarize(
ntaxa = length(unique(taxon_id)))
# Total richness
total_richness <- observation %>%
dplyr::filter(value > 0) %>%
dplyr::select(
.data$datetime,
.data$taxon_id) %>%
dplyr::distinct() %>%
dplyr::mutate(
datetime = dplyr::case_when(
grepl("^(?i)d", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
as.character(),
grepl("^(?i)m", time_window_size) ~ paste0(
.data$datetime %>% lubridate::as_date() %>% lubridate::year(),"-",
.data$datetime %>% lubridate::as_date() %>% lubridate::month(),"-01"),
grepl("^(?i)y", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
lubridate::year() %>%
as.character())) %>%
dplyr::group_by(.data$datetime) %>%
dplyr::summarize(
ntaxa = length(unique(taxon_id))) %>%
dplyr::arrange(.data$datetime) %>%
dplyr::mutate(
location_id = "All sites")
# Format datetime properly based on window size
if(!grepl("^(?i)y", time_window_size)){
richness_by_location$datetime <- richness_by_location$datetime %>%
lubridate::as_date()
total_richness$datetime <- total_richness$datetime %>%
lubridate::as_date()
}else if(grepl("^(?i)y", time_window_size)){
richness_by_location <- richness_by_location %>%
dplyr::mutate(
datetime = paste0(.data$datetime, "-01-01") %>%
lubridate::as_date())
total_richness <- total_richness %>%
dplyr::mutate(
datetime = paste0(.data$datetime, "-01-01") %>%
lubridate::as_date())
}
# Plot
ggplot2::ggplot() +
ggplot2::geom_point(
data = richness_by_location,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = as.factor(.data$location_id),
color = as.factor(.data$location_id)),
alpha = alpha) +
ggplot2::geom_line(
data = richness_by_location,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = as.factor(.data$location_id),
color = as.factor(.data$location_id)),
alpha = alpha) +
ggplot2::geom_point(
data = total_richness,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
fill = ""),
color = "black") +
ggplot2::geom_line(
data = total_richness,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = 1),
color = "black") +
ggplot2::labs(title = "Richness through time", subtitle = id) +
ggplot2::xlab("Year") +
ggplot2::ylab(paste0("Richness")) +
ggplot2::guides(
color = ggplot2::guide_legend(
title = "Site",
label.theme = ggplot2::element_text(size = 6)),
fill = ggplot2::guide_legend(title = "All sites")) +
ggplot2::ylim(c(0, max(total_richness$ntaxa))) +
ggplot2::theme_bw() +
ggplot2::scale_x_date(
date_labels = "%Y",
date_breaks = "1 year",
date_minor_breaks = "1 month") +
ggplot2::theme(axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
}
#' Plot dates and times samples were collected or observations were made
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param color_var (character) Name of column to use to assign colors to the points on the plot
#' @param shape_var (character) Name of column to use to assign shapes to the points on the plot
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_sample_space_time(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(as.numeric(location_id) > 4) %>%
#' plot_sample_space_time()
#' }
#'
#' # Plot the example dataset
#' plot_sample_space_time(ants_L1)
#'
plot_sample_space_time <- function(data,
id = NA_character_,
alpha = 1,
color_var = "package_id",
shape_var = "package_id",
observation = NULL) {
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# # Get observation table and id from data input if possible
# observation <- get_observation_table(data)
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_rank")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
# data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# assing data_long to observation -- legacy of how old fxn was written
observation <- data_long
# Validate inputs
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# summarize data for plot
obs_summary <- observation %>%
# truncate time and just use dates
dplyr::mutate(datetime = .data$datetime %>%
lubridate::as_date() %>%
lubridate::ymd()) %>%
dplyr::select(
all_of(c("event_id","location_id","datetime",
color_var, shape_var))) %>%
dplyr::group_by(
across(
all_of(unique(
c("location_id","datetime",
color_var, shape_var))))) %>%
dplyr::summarize(
`Sampling\nevents` = .data$event_id %>%
unique %>%
length %>%
as.integer)
# Scale font size
uniy <- length(unique(obs_summary$location_id))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# Plot
p <- ggplot2::ggplot() +
ggplot2::geom_point(
data = obs_summary,
ggplot2::aes(x = datetime,
y = location_id,
size = .data$`Sampling\nevents`,
color = .data[[color_var]],
shape = .data[[shape_var]]),
alpha = alpha) +
ggplot2::theme_bw() +
ggplot2::labs(title = "Sample times by location", subtitle = id) +
ggplot2::xlab("Date") +
ggplot2::ylab("Site") +
ggplot2::scale_x_date(
date_labels = "%Y",
date_breaks = "1 year",
date_minor_breaks = "1 month"
) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty),
plot.margin = ggplot2::margin(0.1, 0.25, 0.1, 0.1, "in"),
axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
# do not plot legend for variables that only have one level
if(length(unlist(unique(obs_summary[,color_var]))) == 1){
p <- p + ggplot2::guides(color = "none")}
if(length(unlist(unique(obs_summary[,shape_var]))) == 1){
p <- p + ggplot2::guides(shape = "none")}
if(length(unlist(unique(obs_summary[,"Sampling\nevents"]))) == 1){
p <- p + ggplot2::guides(size = "none")}
# return ggplot object
return(p)
}
#' Plot dates and times samples were taken (DEPRECATED)
#'
#' @description This function has been deprecated. Use \code{plot_sample_space_time()} instead.
#'
#' @param observation (tbl_df, tbl, data.frame) The observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @export
#' @keywords internal
#'
plot_taxa_sample_time <- function(observation, id = NA_character_, alpha = 1) {
.Deprecated("plot_sample_space_time")
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
plot_sample_space_time(data = observation, id = id, alpha = alpha)
}
#' Plot number of unique taxa shared across sites
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_shared_sites(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(as.numeric(location_id) > 4) %>%
#' plot_taxa_shared_sites()
#'
#' # Plot from the observation table directly
#' plot_taxa_shared_sites(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_shared_sites(ants_L1)
#'
plot_taxa_shared_sites <- function(data,
id = NA_character_,
observation = NULL){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
heat_pal_spectral <- grDevices::colorRampPalette(rev( RColorBrewer::brewer.pal(11, "Spectral")))
# Count taxa shared between sites in a site by taxa matrix
shared.species <- function(comm) {
sites <- comm[, 1]
share.mat <- matrix(NA, nrow = length(sites), ncol = length(sites), dimnames = list(sites, sites))
site.pairs <- expand.grid(site1 = sites, site2 = sites)
site.pairs$shared <- NA
for(pair in 1:nrow(site.pairs)) {
site1 <- comm[site.pairs$site1[pair], -1]
site2 <- comm[site.pairs$site2[pair], -1]
site.pairs$shared[pair] <- sum(site1 == 1 & site2 == 1, na.rm = TRUE)
}
return(site.pairs)
}
# aggregate years by cumulative abundances
comm.cumul <- ds$dswide %>%
dplyr::group_by(SITE_ID) %>%
dplyr::select(-OBSERVATION_TYPE, -DATE) %>%
dplyr::summarise_all(sum, na.rm = TRUE)
# Convert sum abundance to presence absence
vals <- comm.cumul %>% dplyr::select(-SITE_ID)
vals[vals > 0] <- 1
comm.cumul <- dplyr::bind_cols(comm.cumul[, 1], vals)
shared.taxa <- shared.species(as.data.frame(comm.cumul))
# scale font size
uniy <- length(unique(shared.taxa$site1))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# Plot
p <- ggplot2::ggplot(shared.taxa, ggplot2::aes(x = site1, y = site2, fill = shared)) +
ggplot2::geom_raster() +
ggplot2::scale_fill_gradientn(colours = heat_pal_spectral(100), name = paste0("Taxa shared")) +
ggplot2::theme_bw() +
ggplot2::labs(title = "Number of taxa shared across sites", subtitle = ds$id) +
ggplot2::xlab("Site") +
ggplot2::ylab("Site") +
ggplot2::theme(
aspect.ratio = 1,
axis.text.x.bottom = ggplot2::element_text(size = txty, angle = 90, hjust = 1, vjust = 0.5),
axis.text.y.left = ggplot2::element_text(size = txty))
if (is.null(txty)) {
p <- p + ggplot2::annotate("text", x = shared.taxa$site1, y = shared.taxa$site2, label = shared.taxa$shared)
} else {
p <- p + ggplot2::annotate("text", x = shared.taxa$site1, y = shared.taxa$site2, label = shared.taxa$shared, size = txty)
}
return(p)
}
#' Plot taxa ranks
#'
#' @description Plot the number of observations that use each taxonomic rank in the dataset.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data(
#' id = "neon.ecocomdp.20120.001.001",
#' site= c('COMO','LECO'),
#' startdate = "2017-06",
#' enddate = "2019-09",
#' check.size = FALSE)
#'
#' # Plot the dataset
#' plot_taxa_rank(dataset)
#'
#' # Plot with facet by location
#' plot_taxa_rank(dataset, facet_var = "location_id")
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(grepl("COMO",location_id)) %>%
#' plot_taxa_rank()
#' }
#'
#' # Plot the example dataset
#' plot_taxa_rank(ants_L1)
#'
plot_taxa_rank <- function(data,
id = NA_character_,
facet_var = NA_character_,
facet_scales = "free_x",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_rank")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# ordered taxon ranks
rank_list_ordered <- c(
"kingdom", "subkingdom", "infrakingdom",
"superphylum", "phylum", "subphylum", "infraphylum",
"superdivision", "division", "subdivision", "infradivision", "parvdivision",
"superclass", "class", "subclass", "infraclass",
"superorder", "order", "suborder", "infraorder",
"section", "subsection",
"superfamily", "family", "subfamily",
"tribe", "subtribe",
"genus", "subgenus", "species", "subspecies",
"variety", "subvariety", "form", "subform",
"stirp", "morph", "aberration", "race") %>%
rev()
# Plot
p <- data_long %>%
dplyr::mutate(
taxon_rank = tolower(taxon_rank) %>%
factor(levels = rank_list_ordered, ordered = TRUE)) %>%
ggplot2::ggplot(
ggplot2::aes(taxon_rank)) +
ggplot2::labs(
title = "Taxa rank frequencies in the observation table",
subtitle = id) +
ggplot2::xlab("Taxon rank") +
ggplot2::ylab(paste0("Number of observations")) +
ggplot2::geom_bar() +
ggplot2::theme_bw() +
ggplot2::coord_flip()
if(!is.na(facet_var)){
p <- p +
ggplot2::facet_wrap(
~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
return(p)
}
#' Plot taxon occurrence frequencies
#'
#' @description Plot taxon occurrence frequences as the number of 'event_id' by 'location_id' combinations in which a taxon is observed.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param min_occurrence (numeric) Minimum number of occurrences allowed for taxa included in the plot.
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param color_var (character) Name of column to use for plot colors.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device.
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_occur_freq(dataset)
#'
#' # Facet by location and color by taxon_rank
#' plot_taxa_occur_freq(
#' data = dataset,
#' facet_var = "location_id",
#' color_var = "taxon_rank")
#'
#' # Color by location and only include taxa with >= 5 occurrences
#' plot_taxa_occur_freq(
#' data = dataset,
#' color_var = "location_id",
#' min_occurrence = 5)
#'
#' # Flatten, filter using a time cutoff, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_occur_freq()
#' }
#' # Plot the example dataset
#' plot_taxa_occur_freq(ants_L1)
#'
plot_taxa_occur_freq <- function(data,
id = NA_character_,
min_occurrence = 0,
facet_var = NA_character_,
color_var = NA_character_,
facet_scales = "free",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_name")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
# data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
# TODO min_occurrence, color_var, facet_var
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# plot title and subtitle text
plot_title = "Taxa occurrence frequencies"
plot_subtitle <- id
if(!is.na(min_occurrence) && min_occurrence > 0) plot_subtitle <- paste0(plot_subtitle,
"\ntaxa with >= ", min_occurrence, " occurrences")
# calculate occurrences
data_working <- data_long %>%
dplyr::filter(.data$value > 0) %>%
dplyr::mutate(occurrence = 1)
col_select_list <- c("event_id","location_id","taxon_name","occurrence",
color_var, facet_var) %>%
stats::na.omit() %>%
unique()
data_working <- data_working[,col_select_list] %>%
dplyr::distinct()
# detemine which taxa meet minimum occurrence threshold in dataset
data_occurrence_total_filtered <- data_working %>%
dplyr::group_by(.data$taxon_name) %>%
dplyr::summarize(
n_occurrences = length(.data$occurrence)) %>%
dplyr::filter(.data$n_occurrences >= min_occurrence)
# calculate occurrence by location_id and event_id for plotting
data_occurrence_by_group <- data_working %>%
dplyr::distinct() %>%
dplyr::filter(
.data$taxon_name %in% data_occurrence_total_filtered$taxon_name) %>%
dplyr::group_by(
dplyr::across(
dplyr::all_of(
unique(stats::na.omit(c("taxon_name", color_var, facet_var)))
)
)
) %>%
dplyr::summarize(
n_occurrences = length(.data$occurrence))
data_occurrence <- data_occurrence_by_group
# Scale font size
uniy <- length(unique(data_occurrence$taxon_name))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# make initial plot object
if(!is.na(color_var)){
p <- data_occurrence %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$n_occurrences),
y = .data$n_occurrences,
color = .data[[color_var]],
fill = .data[[color_var]]))
}else{
p <- data_occurrence %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$n_occurrences),
y = .data$n_occurrences))
}
# add facets if provided
if(!is.na(facet_var)){
p <- p +
ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
# add title, labels and axis formatting
p <- p +
ggplot2::labs(title = plot_title, subtitle = plot_subtitle) +
ggplot2::xlab("Taxon name") +
ggplot2::ylab(paste0("Occurrences\n(unique 'event_id' by 'location_id' occurrences)")) +
ggplot2::geom_col() +
ggplot2::coord_flip() +
ggplot2::scale_x_discrete(limits = rev) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty))
# return plot
return(p)
}
#' Plot mean taxa abundances per 'observation_id'
#'
#' @description Plot taxon abundances averaged across observation records for each taxon. Abundances are reported using the units provided in the dataset. In some cases, these counts are not standardized to sampling effort.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param min_relative_abundance (numeric) Minimum relative abundance allowed for taxa included in the plot; a value between 0 and 1, inclusive.
#' @param trans (character) Define the transform applied to the response variable; "identity" is default, "log1p" is x+1 transform. Built-in transformations include "asn", "atanh", "boxcox", "date", "exp", "hms", "identity", "log", "log10", "log1p", "log2", "logit", "modulus", "probability", "probit", "pseudo_log", "reciprocal", "reverse", "sqrt" and "time".
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param color_var (character) Name of column to use for plot colors.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # plot ecocomDP formatted dataset
#' plot_taxa_abund(dataset)
#'
#' # plot flattened ecocomDP dataset, log(x+1) transform abundances
#' plot_taxa_abund(
#' data = flatten_data(dataset),
#' trans = "log1p")
#'
#' # facet by location color by taxon_rank, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' facet_var = "location_id",
#' color_var = "taxon_rank",
#' trans = "log10")
#'
#' # facet by location, minimum rel. abund = 0.05, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' facet_var = "location_id",
#' min_relative_abundance = 0.05,
#' trans = "log1p")
#'
#' # color by location, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' color_var = "location_id",
#' trans = "log10")
#'
#' # tidy syntax, flatten then filter data by date
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(
#' lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_abund(
#' trans = "log1p",
#' min_relative_abundance = 0.01)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_abund(ants_L1)
#'
plot_taxa_abund <- function(data,
id = NA_character_,
min_relative_abundance = 0,
trans = "identity",
facet_var = NA_character_,
color_var = NA_character_,
facet_scales = "free",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_name")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
flat_data <- data %>% dplyr::distinct()
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
# TODO min_relative_abundance, color_var, facet_var
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# filter based on min relative abund
if(min_relative_abundance > 0){
RA_data_filtered <- flat_data %>%
dplyr::select(.data$taxon_id, .data$value, .data$unit) %>%
dplyr::group_by(.data$taxon_id, .data$unit) %>%
dplyr::summarize(
sum_value = sum(.data$value, na.rm = TRUE)) %>%
dplyr::ungroup() %>%
dplyr::mutate(
total = sum(.data$sum_value),
RA = .data$sum_value / .data$total) %>%
dplyr::filter(.data$RA >= min_relative_abundance)
flat_data <- flat_data %>%
dplyr::filter(.data$taxon_id %in% RA_data_filtered$taxon_id)
}
# plot title and subtitle text
plot_title = "Taxa abundances per 'observation_id'"
plot_subtitle <- id
if(!is.na(min_relative_abundance) && min_relative_abundance > 0) plot_subtitle <- paste0(plot_subtitle,
"\nrel. abundance >= ", min_relative_abundance)
# Scale font size
uniy <- length(unique(flat_data$taxon_name))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# make initial plot object
if(!is.na(color_var)){
p <- flat_data %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$value),
y = .data$value,
color = .data[[color_var]],
fill = .data[[color_var]]))
}else{
p <- flat_data %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$value),
y = .data$value))
}
# add facets if provided
if(!is.na(facet_var)){
if(length(unique(flat_data$unit)) > 1){
p <- p + ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]) +
as.factor(.data$unit),
scales = facet_scales)
}else{
p <- p + ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
}
# add title, labels and axis formatting
p <- p +
ggplot2::labs(title = plot_title, subtitle = plot_subtitle) +
ggplot2::xlab("Taxon name") +
ggplot2::ylab(paste0("Abundance (",paste(unique(flat_data$unit), collapse = " or "), ")")) +
geom_boxplot(alpha = 0.5) + # alpha = transparency
ggplot2::coord_flip() +
ggplot2::scale_x_discrete(limits = rev) +
ggplot2::scale_y_continuous(trans = trans) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty),
axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
return(p)
}
#' Plot sites on US map
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and location tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param color_var (character) Name of column to use to assign colors to the points on the plot
#' @param shape_var (character) Name of column to use to assign shapes to the points on the plot
#' @param labels (logical) Argument to show labels of each US state. Default is TRUE.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 14 columns of the combined observation and location tables.
#'
#' @export
#'
#' @examples
#' \dontrun{
#' library(dplyr)
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_sites(dataset)
#'
#' # Flatten dataset then plot
#' dataset %>%
#' flatten_data() %>%
#' plot_sites()
#'
#' # Download a NEON dataset
#' dataset2 <- read_data(
#' id = "neon.ecocomdp.20120.001.001",
#' site= c('COMO','LECO'),
#' startdate = "2017-06",
#' enddate = "2021-03",
#' token = Sys.getenv("NEON_TOKEN"), # option to use a NEON token
#' check.size = FALSE)
#'
#' # Combine the two datasets and plot. This requires the datasets be first
#' # flattened and then stacked.
#' flattened_data1 <- dataset %>% flatten_data()
#' flattened_data2 <- dataset2 %>% flatten_data()
#' stacked_data <- bind_rows(flattened_data1,flattened_data2)
#' plot_sites(stacked_data)
#' }
#'
#' # Plot the example dataset
#' plot_sites(ants_L1)
#'
plot_sites <- function(
data,
id = NA_character_,
alpha = 1,
labels = TRUE,
color_var = "package_id",
shape_var = "package_id"){
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# check for suggested packages that are required for this function to work
suggs <- c("ggrepel", "usmap", "maps", "sf")
suggsmissing <- !unlist(lapply(suggs, requireNamespace, quietly = TRUE))
if (any(suggsmissing)) {
stop("Packages ", paste(suggs, collapse = ", "), " are required for ",
"running plot_sites(). Packages ",
paste(suggs[suggsmissing], collapse = ", "), " are not installed.",
call. = FALSE)
}
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"location_id","location_name","longitude","latitude")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
flat_data <- data %>% dplyr::distinct()
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
cleaned <- flat_data %>%
dplyr::select(
.data$longitude,
.data$latitude,
.data$location_name,
.data$package_id
) %>%
dplyr::distinct()
cleaned$location_abbrv = substr(cleaned$location_name, 1, 4)
world <- map_data("world")
# Set legend position
legend_position <- "right"
if(length(unlist(unique(flat_data[,color_var]))) == 1 &
length(unlist(unique(flat_data[,shape_var]))) == 1){
legend_position <- "none"
}
# make plot object
p <- ggplot() +
geom_polygon(data = world, aes(x = .data$long, y = .data$lat, group = .data$group), fill = "grey") +
geom_point(data = flat_data, aes(x = .data$longitude, y = .data$latitude,
color = .data[[color_var]],
shape = .data[[shape_var]]),
size = 3) +
labs(x = "Longitude", y = "Latitude", color = color_var) +
ggtitle("US Map with Coordinates") +
theme_bw() +
ggrepel::geom_text_repel(
data = cleaned,
aes(x = .data$longitude,
y = .data$latitude,
label = .data$location_name),
size = 3, max.overlaps = Inf) +
ggplot2::xlab("Longitude") +
ggplot2::ylab("Latitude") +
ggplot2::labs(title = "Map of sites", subtitle = id) +
ggplot2::theme(legend.position = legend_position) +
coord_cartesian(xlim = c(-165, -40), ylim = c(15, 75))
return(p)
}
EDIorg/ecocomDP documentation built on Aug. 22, 2024, 9:34 a.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.
Defines functions plot_sites plot_taxa_abund plot_taxa_occur_freq plot_taxa_rank plot_taxa_shared_sites plot_taxa_sample_time plot_sample_space_time plot_taxa_diversity plot_taxa_accum_time plot_taxa_accum_sites format_for_comm_plots
Documented in plot_sample_space_time plot_sites plot_taxa_abund plot_taxa_accum_sites plot_taxa_accum_time plot_taxa_diversity plot_taxa_occur_freq plot_taxa_rank plot_taxa_sample_time plot_taxa_shared_sites
# Format dataset for community plotting functions
#
# @param observation (tbl_df, tbl, data.frame) The observation table of ecocomDP
# @param id (character) Identifier of dataset to be used in plot subtitles.
#
# @details Downstream plotting functions are based on \href{https://github.com/sokole/ltermetacommunities/tree/master/Group2-explore-data}{LTER Metacommunities code} and use their intermediate data input format.
#
# @return (tbl_df, tbl, data.frame) Tabular data of \code{id} in a format compatible with plotting functions
#
format_for_comm_plots <- function(observation, id) {
id <- id
# Constraints
varname <- unique(observation[c("variable_name", "unit")]) # Can only handle one variable
if (nrow(varname) > 1) {
warning("Input 'observation' has > 1 unique variable_name and unit ",
"combination. Only the first will be used (",
varname$variable_name, ", ", varname$unit, ").", call. = FALSE)
if (is.na(varname$unit[1])) {
observation <- dplyr::filter(observation,
variable_name == varname$variable_name[1] &
is.na(unit))
} else {
observation <- dplyr::filter(observation,
variable_name == varname$variable_name[1] &
unit == varname$unit[1])
}
}
# Methods used by this processing don't require aggregation of non-unique observations, so drop duplicates without warning
dups <- observation %>% dplyr::select(-observation_id, -value, -event_id) %>% duplicated()
if (any(dups)) {
observation <- observation[!dups, ]
}
# Use dates
if (is.character(observation$datetime)) {
warning("Input datetimes are character strings. Accuracy may be improved",
" if inputs are parsed before calling plot_data().", call. = FALSE)
}
observation$datetime <- lubridate::date(observation$datetime)
obs <- observation
taxon_count <- data.frame(
OBSERVATION_TYPE = "TAXON_COUNT",
SITE_ID = obs$location_id,
DATE = obs$datetime,
VARIABLE_NAME = obs$taxon_id,
VARIABLE_UNITS = obs$variable_name,
VALUE = obs$value,
stringsAsFactors = FALSE)
dslong <- taxon_count %>% # long form
dplyr::mutate_at(dplyr::vars(c(VALUE)), as.numeric) %>%
dplyr::mutate_at(dplyr::vars(SITE_ID), as.character)
dswide <- dslong %>% # wide form
dplyr::select(-VARIABLE_UNITS) %>%
tidyr::pivot_wider(
names_from = VARIABLE_NAME,
values_from = VALUE,
values_fn = mean) #average across non-unique observations.
res <- list(
id = id,
dslong = dslong,
dswide = dswide)
return(res)
}
#' Plot taxa accumulation by site accumulation
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_accum_sites(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_accum_sites()
#'
#' # Plot from the observation table directly
#' plot_taxa_accum_sites(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_accum_sites(ants_L1)
#'
plot_taxa_accum_sites <- function(data,
id = NA_character_,
alpha = 1,
observation = NULL) {
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
# Calculate cumulative number of taxa
calc_cuml_taxa_space <- function(ex) {
taxa.s.list <- list()
sites <- unique(ex$SITE_ID)
for(t in 1:length(unique(ex$SITE_ID))){ # unique taxa found in each year
tmp.dat <- subset(ex, ex$SITE_ID == sites[t])
tmp.dat.pres <- subset(tmp.dat, tmp.dat$VALUE > 0)
taxa.s.list[[t]] <- unique(tmp.dat.pres$VARIABLE_NAME)
}
cuml.taxa.space <- list() # cumulative list of taxa over space
cuml.taxa.space[[1]] <- taxa.s.list[[1]]
if (length(unique(ex$SITE_ID)) > 1) {
for(t in 2:length(unique(ex$SITE_ID))){ # list cumulative taxa, with duplicates
cuml.taxa.space[[t]] <- c(cuml.taxa.space[[t - 1]], taxa.s.list[[t]])
}
}
cuml.taxa.space <- lapply(cuml.taxa.space, function(x) {unique(x)}) # rm duplicates
cuml.no.taxa.space <- data.frame("site" = unique(ex$SITE_ID)) # total unique taxa over space
cuml.no.taxa.space$no.taxa <- unlist(lapply(cuml.taxa.space, function(x) {length(x)}))
return(cuml.no.taxa.space)
}
comm.dat <- ds$dslong %>% arrange(SITE_ID) # order by site
no.taxa.space <- calc_cuml_taxa_space(comm.dat)
no.taxa.space$no.site <- as.numeric(rownames(no.taxa.space))
# Plot
p <- ggplot2::ggplot(data = no.taxa.space, ggplot2::aes(x = no.site, y = no.taxa)) +
ggplot2::geom_point() +
ggplot2::geom_line() +
ggplot2::labs(title = "Taxa accumulation by site accumulation", subtitle = ds$id) +
ggplot2::xlab("Cumulative number of sites") +
ggplot2::ylab(paste0("Cumulative number of taxa")) +
ggplot2::theme_bw()
return(p)
}
#' Plot taxa accumulation through time
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_accum_time(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_accum_time()
#'
#' # Plot from the observation table directly
#' plot_taxa_accum_time(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_accum_time(ants_L1)
#'
plot_taxa_accum_time <- function(data,
id = NA_character_,
alpha = 1,
observation = NULL){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
# Calculate cumulative number of taxa
cuml.taxa.fun <- function(ex){
taxa.t.list <- list()
dates <- unique(ex$DATE)
for(t in 1:length(unique(ex$DATE))) { # unique taxa found in each year
tmp.dat <- subset(ex, ex$DATE == dates[t])
tmp.dat.pres <- subset(tmp.dat, tmp.dat$VALUE > 0)
taxa.t.list[[t]] <- unique(tmp.dat.pres$VARIABLE_NAME)
}
cuml.taxa <- list() # cumulative taxa through time
cuml.taxa[[1]] <- taxa.t.list[[1]]
if (length(unique(ex$DATE)) > 1) { # create list of the cumulative taxa, with duplicates
for(t in 2:length(unique(ex$DATE))){
cuml.taxa[[t]] <- c(cuml.taxa[[t - 1]], taxa.t.list[[t]])
}
}
cuml.taxa <- lapply(cuml.taxa, function(x){unique(x)}) # rm duplicates
cuml.no.taxa <- data.frame("year" = unique(ex$DATE), stringsAsFactors = FALSE) # number of unique taxa through time
cuml.no.taxa$no.taxa <- unlist(lapply(cuml.taxa, function(x){length(x)}))
return(cuml.no.taxa)
}
cuml.taxa.all.sites <- cuml.taxa.fun(
ex = ds$dslong %>% dplyr::arrange(DATE)) # taxa accumulation across all sites pooled together
# Plot
p <- ggplot2::ggplot() +
ggplot2::geom_point(
data = cuml.taxa.all.sites,
ggplot2::aes(x = year, y = no.taxa, fill = "")) +
ggplot2::geom_line(
data = cuml.taxa.all.sites,
ggplot2::aes(x = year, y = no.taxa)) +
ggplot2::labs(title = "Accumulation of taxa through time", subtitle = ds$id) +
ggplot2::xlab("Year") +
ggplot2::ylab(paste0("Cumulative number of taxa")) +
ggplot2::scale_x_date(date_labels = "%Y", date_breaks = "1 year", date_minor_breaks = "1 month") +
ggplot2::guides(fill = "none") +
ggplot2::ylim(c(0, max(cuml.taxa.all.sites$no.taxa))) +
ggplot2::theme_bw() +
theme(axis.text.x = element_text(angle = 45, hjust=1))
return(p)
}
#' Plot diversity (taxa richness) through time
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param time_window_size (character) Define the time window over which to aggregate observations for calculating richness. Can be: "day" or "year"
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_diversity(dataset)
#'
#' # Plot the dataset with observations aggregated by year
#' plot_taxa_diversity(dataset, time_window_size = "year")
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(
#' lubridate::as_date(datetime) > "2007-01-01") %>%
#' plot_taxa_diversity()
#'
#' # Plot from the observation table directly
#' plot_taxa_diversity(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_diversity(ants_L1)
#'
plot_taxa_diversity <- function(data,
id = NA_character_,
time_window_size = "day",
observation = NULL,
alpha = 1){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# Richness by location id
richness_by_location <- observation %>%
dplyr::filter(value > 0) %>%
dplyr::select(
.data$location_id,
.data$datetime,
.data$taxon_id) %>%
dplyr::distinct() %>%
dplyr::mutate(
datetime = dplyr::case_when(
grepl("^(?i)d", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
as.character(),
grepl("^(?i)m", time_window_size) ~ paste0(
.data$datetime %>% lubridate::as_date() %>% lubridate::year(),"-",
.data$datetime %>% lubridate::as_date() %>% lubridate::month(),"-01"),
grepl("^(?i)y", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
lubridate::year() %>%
as.character())) %>%
dplyr::group_by(.data$location_id, .data$datetime) %>%
dplyr::summarize(
ntaxa = length(unique(taxon_id)))
# Total richness
total_richness <- observation %>%
dplyr::filter(value > 0) %>%
dplyr::select(
.data$datetime,
.data$taxon_id) %>%
dplyr::distinct() %>%
dplyr::mutate(
datetime = dplyr::case_when(
grepl("^(?i)d", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
as.character(),
grepl("^(?i)m", time_window_size) ~ paste0(
.data$datetime %>% lubridate::as_date() %>% lubridate::year(),"-",
.data$datetime %>% lubridate::as_date() %>% lubridate::month(),"-01"),
grepl("^(?i)y", time_window_size) ~ .data$datetime %>%
lubridate::as_date() %>%
lubridate::year() %>%
as.character())) %>%
dplyr::group_by(.data$datetime) %>%
dplyr::summarize(
ntaxa = length(unique(taxon_id))) %>%
dplyr::arrange(.data$datetime) %>%
dplyr::mutate(
location_id = "All sites")
# Format datetime properly based on window size
if(!grepl("^(?i)y", time_window_size)){
richness_by_location$datetime <- richness_by_location$datetime %>%
lubridate::as_date()
total_richness$datetime <- total_richness$datetime %>%
lubridate::as_date()
}else if(grepl("^(?i)y", time_window_size)){
richness_by_location <- richness_by_location %>%
dplyr::mutate(
datetime = paste0(.data$datetime, "-01-01") %>%
lubridate::as_date())
total_richness <- total_richness %>%
dplyr::mutate(
datetime = paste0(.data$datetime, "-01-01") %>%
lubridate::as_date())
}
# Plot
ggplot2::ggplot() +
ggplot2::geom_point(
data = richness_by_location,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = as.factor(.data$location_id),
color = as.factor(.data$location_id)),
alpha = alpha) +
ggplot2::geom_line(
data = richness_by_location,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = as.factor(.data$location_id),
color = as.factor(.data$location_id)),
alpha = alpha) +
ggplot2::geom_point(
data = total_richness,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
fill = ""),
color = "black") +
ggplot2::geom_line(
data = total_richness,
ggplot2::aes(x = .data$datetime,
y = .data$ntaxa,
group = 1),
color = "black") +
ggplot2::labs(title = "Richness through time", subtitle = id) +
ggplot2::xlab("Year") +
ggplot2::ylab(paste0("Richness")) +
ggplot2::guides(
color = ggplot2::guide_legend(
title = "Site",
label.theme = ggplot2::element_text(size = 6)),
fill = ggplot2::guide_legend(title = "All sites")) +
ggplot2::ylim(c(0, max(total_richness$ntaxa))) +
ggplot2::theme_bw() +
ggplot2::scale_x_date(
date_labels = "%Y",
date_breaks = "1 year",
date_minor_breaks = "1 month") +
ggplot2::theme(axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
}
#' Plot dates and times samples were collected or observations were made
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param color_var (character) Name of column to use to assign colors to the points on the plot
#' @param shape_var (character) Name of column to use to assign shapes to the points on the plot
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_sample_space_time(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(as.numeric(location_id) > 4) %>%
#' plot_sample_space_time()
#' }
#'
#' # Plot the example dataset
#' plot_sample_space_time(ants_L1)
#'
plot_sample_space_time <- function(data,
id = NA_character_,
alpha = 1,
color_var = "package_id",
shape_var = "package_id",
observation = NULL) {
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# # Get observation table and id from data input if possible
# observation <- get_observation_table(data)
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_rank")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
# data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# assing data_long to observation -- legacy of how old fxn was written
observation <- data_long
# Validate inputs
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# summarize data for plot
obs_summary <- observation %>%
# truncate time and just use dates
dplyr::mutate(datetime = .data$datetime %>%
lubridate::as_date() %>%
lubridate::ymd()) %>%
dplyr::select(
all_of(c("event_id","location_id","datetime",
color_var, shape_var))) %>%
dplyr::group_by(
across(
all_of(unique(
c("location_id","datetime",
color_var, shape_var))))) %>%
dplyr::summarize(
`Sampling\nevents` = .data$event_id %>%
unique %>%
length %>%
as.integer)
# Scale font size
uniy <- length(unique(obs_summary$location_id))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# Plot
p <- ggplot2::ggplot() +
ggplot2::geom_point(
data = obs_summary,
ggplot2::aes(x = datetime,
y = location_id,
size = .data$`Sampling\nevents`,
color = .data[[color_var]],
shape = .data[[shape_var]]),
alpha = alpha) +
ggplot2::theme_bw() +
ggplot2::labs(title = "Sample times by location", subtitle = id) +
ggplot2::xlab("Date") +
ggplot2::ylab("Site") +
ggplot2::scale_x_date(
date_labels = "%Y",
date_breaks = "1 year",
date_minor_breaks = "1 month"
) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty),
plot.margin = ggplot2::margin(0.1, 0.25, 0.1, 0.1, "in"),
axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
# do not plot legend for variables that only have one level
if(length(unlist(unique(obs_summary[,color_var]))) == 1){
p <- p + ggplot2::guides(color = "none")}
if(length(unlist(unique(obs_summary[,shape_var]))) == 1){
p <- p + ggplot2::guides(shape = "none")}
if(length(unlist(unique(obs_summary[,"Sampling\nevents"]))) == 1){
p <- p + ggplot2::guides(size = "none")}
# return ggplot object
return(p)
}
#' Plot dates and times samples were taken (DEPRECATED)
#'
#' @description This function has been deprecated. Use \code{plot_sample_space_time()} instead.
#'
#' @param observation (tbl_df, tbl, data.frame) The observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @export
#' @keywords internal
#'
plot_taxa_sample_time <- function(observation, id = NA_character_, alpha = 1) {
.Deprecated("plot_sample_space_time")
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
plot_sample_space_time(data = observation, id = id, alpha = alpha)
}
#' Plot number of unique taxa shared across sites
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation table, or a flat table containing columns of the observation table.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param observation (tbl_df, tbl, data.frame) DEPRECATED: Use \code{data} instead.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 9 columns of the observation table.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_shared_sites(dataset)
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(as.numeric(location_id) > 4) %>%
#' plot_taxa_shared_sites()
#'
#' # Plot from the observation table directly
#' plot_taxa_shared_sites(dataset$tables$observation)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_shared_sites(ants_L1)
#'
plot_taxa_shared_sites <- function(data,
id = NA_character_,
observation = NULL){
# Warn about deprecated observation parameter (remove after 2022-10-18)
if (!is.null(observation)) {
data <- observation
warning('Input argument "observation" is deprecated, use "data" instead',
call. = FALSE)
# get data package id from dataset object
if (is.na(id)) {
id <- get_id(data)
}
}
# Get observation table and id from data input if possible
observation <- get_observation_table(data)
# if id is still missing, get it from observation table
if (is.na(id)){
id <- paste(unique(observation$package_id), collapse = "|")
}
# Validate inputs and construct intermediate format for plotting
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
ds <- format_for_comm_plots(observation, id)
heat_pal_spectral <- grDevices::colorRampPalette(rev( RColorBrewer::brewer.pal(11, "Spectral")))
# Count taxa shared between sites in a site by taxa matrix
shared.species <- function(comm) {
sites <- comm[, 1]
share.mat <- matrix(NA, nrow = length(sites), ncol = length(sites), dimnames = list(sites, sites))
site.pairs <- expand.grid(site1 = sites, site2 = sites)
site.pairs$shared <- NA
for(pair in 1:nrow(site.pairs)) {
site1 <- comm[site.pairs$site1[pair], -1]
site2 <- comm[site.pairs$site2[pair], -1]
site.pairs$shared[pair] <- sum(site1 == 1 & site2 == 1, na.rm = TRUE)
}
return(site.pairs)
}
# aggregate years by cumulative abundances
comm.cumul <- ds$dswide %>%
dplyr::group_by(SITE_ID) %>%
dplyr::select(-OBSERVATION_TYPE, -DATE) %>%
dplyr::summarise_all(sum, na.rm = TRUE)
# Convert sum abundance to presence absence
vals <- comm.cumul %>% dplyr::select(-SITE_ID)
vals[vals > 0] <- 1
comm.cumul <- dplyr::bind_cols(comm.cumul[, 1], vals)
shared.taxa <- shared.species(as.data.frame(comm.cumul))
# scale font size
uniy <- length(unique(shared.taxa$site1))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# Plot
p <- ggplot2::ggplot(shared.taxa, ggplot2::aes(x = site1, y = site2, fill = shared)) +
ggplot2::geom_raster() +
ggplot2::scale_fill_gradientn(colours = heat_pal_spectral(100), name = paste0("Taxa shared")) +
ggplot2::theme_bw() +
ggplot2::labs(title = "Number of taxa shared across sites", subtitle = ds$id) +
ggplot2::xlab("Site") +
ggplot2::ylab("Site") +
ggplot2::theme(
aspect.ratio = 1,
axis.text.x.bottom = ggplot2::element_text(size = txty, angle = 90, hjust = 1, vjust = 0.5),
axis.text.y.left = ggplot2::element_text(size = txty))
if (is.null(txty)) {
p <- p + ggplot2::annotate("text", x = shared.taxa$site1, y = shared.taxa$site2, label = shared.taxa$shared)
} else {
p <- p + ggplot2::annotate("text", x = shared.taxa$site1, y = shared.taxa$site2, label = shared.taxa$shared, size = txty)
}
return(p)
}
#' Plot taxa ranks
#'
#' @description Plot the number of observations that use each taxonomic rank in the dataset.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data(
#' id = "neon.ecocomdp.20120.001.001",
#' site= c('COMO','LECO'),
#' startdate = "2017-06",
#' enddate = "2019-09",
#' check.size = FALSE)
#'
#' # Plot the dataset
#' plot_taxa_rank(dataset)
#'
#' # Plot with facet by location
#' plot_taxa_rank(dataset, facet_var = "location_id")
#'
#' # Flatten the dataset, manipulate, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' dplyr::filter(grepl("COMO",location_id)) %>%
#' plot_taxa_rank()
#' }
#'
#' # Plot the example dataset
#' plot_taxa_rank(ants_L1)
#'
plot_taxa_rank <- function(data,
id = NA_character_,
facet_var = NA_character_,
facet_scales = "free_x",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_rank")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# ordered taxon ranks
rank_list_ordered <- c(
"kingdom", "subkingdom", "infrakingdom",
"superphylum", "phylum", "subphylum", "infraphylum",
"superdivision", "division", "subdivision", "infradivision", "parvdivision",
"superclass", "class", "subclass", "infraclass",
"superorder", "order", "suborder", "infraorder",
"section", "subsection",
"superfamily", "family", "subfamily",
"tribe", "subtribe",
"genus", "subgenus", "species", "subspecies",
"variety", "subvariety", "form", "subform",
"stirp", "morph", "aberration", "race") %>%
rev()
# Plot
p <- data_long %>%
dplyr::mutate(
taxon_rank = tolower(taxon_rank) %>%
factor(levels = rank_list_ordered, ordered = TRUE)) %>%
ggplot2::ggplot(
ggplot2::aes(taxon_rank)) +
ggplot2::labs(
title = "Taxa rank frequencies in the observation table",
subtitle = id) +
ggplot2::xlab("Taxon rank") +
ggplot2::ylab(paste0("Number of observations")) +
ggplot2::geom_bar() +
ggplot2::theme_bw() +
ggplot2::coord_flip()
if(!is.na(facet_var)){
p <- p +
ggplot2::facet_wrap(
~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
return(p)
}
#' Plot taxon occurrence frequencies
#'
#' @description Plot taxon occurrence frequences as the number of 'event_id' by 'location_id' combinations in which a taxon is observed.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param min_occurrence (numeric) Minimum number of occurrences allowed for taxa included in the plot.
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param color_var (character) Name of column to use for plot colors.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device.
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_taxa_occur_freq(dataset)
#'
#' # Facet by location and color by taxon_rank
#' plot_taxa_occur_freq(
#' data = dataset,
#' facet_var = "location_id",
#' color_var = "taxon_rank")
#'
#' # Color by location and only include taxa with >= 5 occurrences
#' plot_taxa_occur_freq(
#' data = dataset,
#' color_var = "location_id",
#' min_occurrence = 5)
#'
#' # Flatten, filter using a time cutoff, then plot
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_occur_freq()
#' }
#' # Plot the example dataset
#' plot_taxa_occur_freq(ants_L1)
#'
plot_taxa_occur_freq <- function(data,
id = NA_character_,
min_occurrence = 0,
facet_var = NA_character_,
color_var = NA_character_,
facet_scales = "free",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_name")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
data_long <- data %>% dplyr::distinct()
# data_long <- data_long[,req_col_names]
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
data_long <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
data_long <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
# TODO min_occurrence, color_var, facet_var
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# plot title and subtitle text
plot_title = "Taxa occurrence frequencies"
plot_subtitle <- id
if(!is.na(min_occurrence) && min_occurrence > 0) plot_subtitle <- paste0(plot_subtitle,
"\ntaxa with >= ", min_occurrence, " occurrences")
# calculate occurrences
data_working <- data_long %>%
dplyr::filter(.data$value > 0) %>%
dplyr::mutate(occurrence = 1)
col_select_list <- c("event_id","location_id","taxon_name","occurrence",
color_var, facet_var) %>%
stats::na.omit() %>%
unique()
data_working <- data_working[,col_select_list] %>%
dplyr::distinct()
# detemine which taxa meet minimum occurrence threshold in dataset
data_occurrence_total_filtered <- data_working %>%
dplyr::group_by(.data$taxon_name) %>%
dplyr::summarize(
n_occurrences = length(.data$occurrence)) %>%
dplyr::filter(.data$n_occurrences >= min_occurrence)
# calculate occurrence by location_id and event_id for plotting
data_occurrence_by_group <- data_working %>%
dplyr::distinct() %>%
dplyr::filter(
.data$taxon_name %in% data_occurrence_total_filtered$taxon_name) %>%
dplyr::group_by(
dplyr::across(
dplyr::all_of(
unique(stats::na.omit(c("taxon_name", color_var, facet_var)))
)
)
) %>%
dplyr::summarize(
n_occurrences = length(.data$occurrence))
data_occurrence <- data_occurrence_by_group
# Scale font size
uniy <- length(unique(data_occurrence$taxon_name))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# make initial plot object
if(!is.na(color_var)){
p <- data_occurrence %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$n_occurrences),
y = .data$n_occurrences,
color = .data[[color_var]],
fill = .data[[color_var]]))
}else{
p <- data_occurrence %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$n_occurrences),
y = .data$n_occurrences))
}
# add facets if provided
if(!is.na(facet_var)){
p <- p +
ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
# add title, labels and axis formatting
p <- p +
ggplot2::labs(title = plot_title, subtitle = plot_subtitle) +
ggplot2::xlab("Taxon name") +
ggplot2::ylab(paste0("Occurrences\n(unique 'event_id' by 'location_id' occurrences)")) +
ggplot2::geom_col() +
ggplot2::coord_flip() +
ggplot2::scale_x_discrete(limits = rev) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty))
# return plot
return(p)
}
#' Plot mean taxa abundances per 'observation_id'
#'
#' @description Plot taxon abundances averaged across observation records for each taxon. Abundances are reported using the units provided in the dataset. In some cases, these counts are not standardized to sampling effort.
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and taxon tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param min_relative_abundance (numeric) Minimum relative abundance allowed for taxa included in the plot; a value between 0 and 1, inclusive.
#' @param trans (character) Define the transform applied to the response variable; "identity" is default, "log1p" is x+1 transform. Built-in transformations include "asn", "atanh", "boxcox", "date", "exp", "hms", "identity", "log", "log10", "log1p", "log2", "logit", "modulus", "probability", "probit", "pseudo_log", "reciprocal", "reverse", "sqrt" and "time".
#' @param facet_var (character) Name of column to use for faceting. Must be a column of the observation or taxon table.
#' @param facet_scales (character) Should scales be free ("free", default value), fixed ("fixed"), or free in one dimension ("free_x", "free_y")?
#' @param color_var (character) Name of column to use for plot colors.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 13 columns of the combined observation and taxon tables.
#'
#' @import dplyr
#' @import ggplot2
#' @import tidyr
#'
#' @export
#'
#' @examples
#' \dontrun{
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # plot ecocomDP formatted dataset
#' plot_taxa_abund(dataset)
#'
#' # plot flattened ecocomDP dataset, log(x+1) transform abundances
#' plot_taxa_abund(
#' data = flatten_data(dataset),
#' trans = "log1p")
#'
#' # facet by location color by taxon_rank, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' facet_var = "location_id",
#' color_var = "taxon_rank",
#' trans = "log10")
#'
#' # facet by location, minimum rel. abund = 0.05, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' facet_var = "location_id",
#' min_relative_abundance = 0.05,
#' trans = "log1p")
#'
#' # color by location, log 10 transform
#' plot_taxa_abund(
#' data = dataset,
#' color_var = "location_id",
#' trans = "log10")
#'
#' # tidy syntax, flatten then filter data by date
#' dataset %>%
#' flatten_data() %>%
#' dplyr::filter(
#' lubridate::as_date(datetime) > "2003-07-01") %>%
#' plot_taxa_abund(
#' trans = "log1p",
#' min_relative_abundance = 0.01)
#' }
#'
#' # Plot the example dataset
#' plot_taxa_abund(ants_L1)
#'
plot_taxa_abund <- function(data,
id = NA_character_,
min_relative_abundance = 0,
trans = "identity",
facet_var = NA_character_,
color_var = NA_character_,
facet_scales = "free",
alpha = 1) {
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"taxon_name")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
flat_data <- data %>% dplyr::distinct()
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
# Validate inputs
# TODO min_relative_abundance, color_var, facet_var
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# filter based on min relative abund
if(min_relative_abundance > 0){
RA_data_filtered <- flat_data %>%
dplyr::select(.data$taxon_id, .data$value, .data$unit) %>%
dplyr::group_by(.data$taxon_id, .data$unit) %>%
dplyr::summarize(
sum_value = sum(.data$value, na.rm = TRUE)) %>%
dplyr::ungroup() %>%
dplyr::mutate(
total = sum(.data$sum_value),
RA = .data$sum_value / .data$total) %>%
dplyr::filter(.data$RA >= min_relative_abundance)
flat_data <- flat_data %>%
dplyr::filter(.data$taxon_id %in% RA_data_filtered$taxon_id)
}
# plot title and subtitle text
plot_title = "Taxa abundances per 'observation_id'"
plot_subtitle <- id
if(!is.na(min_relative_abundance) && min_relative_abundance > 0) plot_subtitle <- paste0(plot_subtitle,
"\nrel. abundance >= ", min_relative_abundance)
# Scale font size
uniy <- length(unique(flat_data$taxon_name))
if (uniy < 30) {
txty <- NULL
} else if (uniy < 60) {
txty <- 6
} else if (uniy >= 60) {
txty <- 4
}
# make initial plot object
if(!is.na(color_var)){
p <- flat_data %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$value),
y = .data$value,
color = .data[[color_var]],
fill = .data[[color_var]]))
}else{
p <- flat_data %>%
ggplot2::ggplot(
ggplot2::aes(
x = stats::reorder(.data$taxon_name, -.data$value),
y = .data$value))
}
# add facets if provided
if(!is.na(facet_var)){
if(length(unique(flat_data$unit)) > 1){
p <- p + ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]) +
as.factor(.data$unit),
scales = facet_scales)
}else{
p <- p + ggplot2::facet_wrap(~ as.factor(.data[[facet_var]]),
scales = facet_scales)
}
}
# add title, labels and axis formatting
p <- p +
ggplot2::labs(title = plot_title, subtitle = plot_subtitle) +
ggplot2::xlab("Taxon name") +
ggplot2::ylab(paste0("Abundance (",paste(unique(flat_data$unit), collapse = " or "), ")")) +
geom_boxplot(alpha = 0.5) + # alpha = transparency
ggplot2::coord_flip() +
ggplot2::scale_x_discrete(limits = rev) +
ggplot2::scale_y_continuous(trans = trans) +
ggplot2::theme_bw() +
ggplot2::theme(
axis.text.y.left = ggplot2::element_text(size = txty),
axis.text.x = ggplot2::element_text(angle = 45, hjust = 1))
return(p)
}
#' Plot sites on US map
#'
#' @param data (list or tbl_df, tbl, data.frame) The dataset object returned by \code{read_data()}, a named list of tables containing the observation and taxon tables, or a flat table containing columns of the observation and location tables.
#' @param id (character) Identifier of dataset to be used in plot subtitles. Is automatically assigned when \code{data} is a dataset object containing the \code{id} field, or is a table containing the package_id column.
#' @param alpha (numeric) Alpha-transparency scale of data points. Useful when many data points overlap. Allowed values are between 0 and 1, where 1 is 100\% opaque. Default is 1.
#' @param color_var (character) Name of column to use to assign colors to the points on the plot
#' @param shape_var (character) Name of column to use to assign shapes to the points on the plot
#' @param labels (logical) Argument to show labels of each US state. Default is TRUE.
#'
#' @return (gg, ggplot) A gg, ggplot object if assigned to a variable, otherwise a plot to your active graphics device
#'
#' @details The \code{data} parameter accepts a range of input types but ultimately requires the 14 columns of the combined observation and location tables.
#'
#' @export
#'
#' @examples
#' \dontrun{
#' library(dplyr)
#' # Read a dataset of interest
#' dataset <- read_data("edi.193.5")
#'
#' # Plot the dataset
#' plot_sites(dataset)
#'
#' # Flatten dataset then plot
#' dataset %>%
#' flatten_data() %>%
#' plot_sites()
#'
#' # Download a NEON dataset
#' dataset2 <- read_data(
#' id = "neon.ecocomdp.20120.001.001",
#' site= c('COMO','LECO'),
#' startdate = "2017-06",
#' enddate = "2021-03",
#' token = Sys.getenv("NEON_TOKEN"), # option to use a NEON token
#' check.size = FALSE)
#'
#' # Combine the two datasets and plot. This requires the datasets be first
#' # flattened and then stacked.
#' flattened_data1 <- dataset %>% flatten_data()
#' flattened_data2 <- dataset2 %>% flatten_data()
#' stacked_data <- bind_rows(flattened_data1,flattened_data2)
#' plot_sites(stacked_data)
#' }
#'
#' # Plot the example dataset
#' plot_sites(ants_L1)
#'
plot_sites <- function(
data,
id = NA_character_,
alpha = 1,
labels = TRUE,
color_var = "package_id",
shape_var = "package_id"){
validate_arguments(fun.name = "plot", fun.args = as.list(environment()))
# check for suggested packages that are required for this function to work
suggs <- c("ggrepel", "usmap", "maps", "sf")
suggsmissing <- !unlist(lapply(suggs, requireNamespace, quietly = TRUE))
if (any(suggsmissing)) {
stop("Packages ", paste(suggs, collapse = ", "), " are required for ",
"running plot_sites(). Packages ",
paste(suggs[suggsmissing], collapse = ", "), " are not installed.",
call. = FALSE)
}
# TODO: Convert this unreadable block of code to a function like get_observation_table()
# TODO: Call get_id()
# required col names in flat data
req_col_names <- c("observation_id","event_id","package_id","location_id",
"datetime","taxon_id","variable_name","value",
"location_id","location_name","longitude","latitude")
# detect data type, extract observation table
data_type <- detect_data_type(data)
if(data_type == "table" && all(req_col_names %in% names(data))){
flat_data <- data %>% dplyr::distinct()
if(is.na(id)) id <- paste(unique(data$package_id), collapse = " | ")
}else if(data_type == "table" && !all(req_col_names %in% names(data))){
stop("please provide a valid ecocomDP dataset or a table that includes the columns present in the ecocomDP 'observation' and 'taxon' tables")
}else if(data_type == "dataset_old"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- names(data)
}else if(data_type == "dataset"){
flat_data <- flatten_data(data)
if(is.na(id)) id <- data$id
}else{
stop("No plotting method currently implemented for this data format")
}
cleaned <- flat_data %>%
dplyr::select(
.data$longitude,
.data$latitude,
.data$location_name,
.data$package_id
) %>%
dplyr::distinct()
cleaned$location_abbrv = substr(cleaned$location_name, 1, 4)
world <- map_data("world")
# Set legend position
legend_position <- "right"
if(length(unlist(unique(flat_data[,color_var]))) == 1 &
length(unlist(unique(flat_data[,shape_var]))) == 1){
legend_position <- "none"
}
# make plot object
p <- ggplot() +
geom_polygon(data = world, aes(x = .data$long, y = .data$lat, group = .data$group), fill = "grey") +
geom_point(data = flat_data, aes(x = .data$longitude, y = .data$latitude,
color = .data[[color_var]],
shape = .data[[shape_var]]),
size = 3) +
labs(x = "Longitude", y = "Latitude", color = color_var) +
ggtitle("US Map with Coordinates") +
theme_bw() +
ggrepel::geom_text_repel(
data = cleaned,
aes(x = .data$longitude,
y = .data$latitude,
label = .data$location_name),
size = 3, max.overlaps = Inf) +
ggplot2::xlab("Longitude") +
ggplot2::ylab("Latitude") +
ggplot2::labs(title = "Map of sites", subtitle = id) +
ggplot2::theme(legend.position = legend_position) +
coord_cartesian(xlim = c(-165, -40), ylim = c(15, 75))
return(p)
}
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.