R/plotting.R
In VCCRI/CardiacProfileR: Generate Heart Rate Dynamic Profiles from TCX Data
# plotting.R
# functions that can plot profiles
####################
# helper functions #
####################
#' Given an index returns a plotly color
#' @description returns a plotly color to be used in plotting functions
#'
#' @param i Index of color
#' @return Color to be used in plotting function
#' @family plotting functions
get_color <- function(i) {
colors <- list(
'#1f77b4', # muted blue
'#ff7f0e', # safety orange
'#2ca02c', # cooked asparagus green
'#d62728', # brick red
'#9467bd', # muted purple
'#8c564b', # chestnut brown
'#e377c2', # raspberry yogurt pink
'#7f7f7f', # middle gray
'#bcbd22', # curry yellow-green
'#17becf' # blue-teal
)
nc = length(colors) # number of colors
return(colors[[i %% nc + 1]])
}
#' Get ymin ymax from plot
#' @description returns the minimum and maximum y values from a plotly function
#'
#' @param p Plotly plot
#' @return list with:
#' \itemize{
#' \item min : minimum y value
#' \item max : maximum y value
#' }
#' @family plotting functions
min_max_plot <- function(p) {
yvals <- unlist(lapply(p$x$attrs, function(x) {
if ("y" %in% names(x)) return(x$y)
if ("ymin" %in% names(x)) return(c(x$ymin, x$ymax))
}))
yvals <- yvals[!is.na(yvals)]
bounds <- list("min" = floor(min(yvals))-1, "max" = ceiling(max(yvals))+1)
return(bounds)
}
#####################
# plotting function #
#####################
#' Plot a single profile
#' @description plot single profiles individually or add to existing plot
#'
#' @param profile Profile or aggregate profile to be plotted.
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param label Label of the profile being plotted
#' @param title Title of the plot
#' @param color Color of the profile added to the plot.
#' @param p Plot to be added to, if none provided a new plot will be created.
#' @param sections Style used to plot sections of the profile.
#' See \code{\link{get_profile}} for description of profile sections.
#'
#' The style options are:
#' \itemize{
#' \item "shade" : shade the background of the different sections different colors
#' \item "lines" : add lines to separate the different sections
#' }
#' @param sec_opacity The opacity of the sections of the profile added to the plot.
#' @param ribbon_opacity The opacity of standard deviation ribbon when plotting aggregate profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @return Plotly object.
#' @examples
#' profile1 <- get_profile(df, 500, 500, 500)
#' p <- plot_profile(profile1, "profile 1", title = "Great Profiles")
#' profile2 <- get_profile(df, 200, 300, 500)
#' plot_profile(profile2, "profile 2", p = p)
#'
#' profiles <- get_profiles(dfs, 500, 500, 500)
#' agg_profile <- aggregate_profiles(profiles)
#' plot_profile(agg_profile, "aggregated profile", ribbon_opacity = 0.2)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_profile <- function(profile, label, title = NULL, color = get_color(0),
p = plot_ly(), sections = "shade",
sec_opacity = 0.3, ribbon_opactiy = 0.4) {
# check if it's an aggregate profile or not
if (Reduce("&", c("mean", "sd") %in% names(profile))) {
is_aggregate = TRUE
} else {
is_aggregate = FALSE
}
# plot data
profile <- data.frame(profile)
xval = c(1:nrow(profile))
if (is_aggregate) {
# plot mean line
p <- add_lines(p, x = xval, y = profile$mean,
text = paste('sd: ', profile$sd),
line = list(color = color),
legendgroup = label, name = label)
# add sd ribbon
p <- add_ribbons(p, x = xval,
ymin = profile$mean - profile$sd,
ymax = profile$mean + profile$sd,
line = list(color = color),
opacity = ribbon_opactiy, hoverinfo = "none",
legendgroup = label, name = label,
inherit = TRUE, showlegend = FALSE)
} else {
# plot individual profile
p <- add_lines(p, x = xval, y = profile$profile, name = label)
}
p <- layout(p,
xaxis = list(title = "Sections", showticklabels = FALSE),
yaxis = list(title = "BPM"),
title = title)
# determine type of sectioning
sections <- if (is.null(sections)) 'none' else sections
section_option <- list("lines" = FALSE, "shade" = TRUE)[[sections]]
# add sections (either as lines or shading)
if (!is.null(section_option)) {
p <- add_sections(p, shade = section_option, opacity = sec_opacity, scale = nrow(profile)/3)
}
return(p)
}
#' Plot a many individual profiles
#' @description plot many profiles or aggregate profiles on one plot
#'
#' @param profiles A list of profiles or aggregate profiles to be plotted.
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param labels A list of labels of the profiles being plotted.
#' @param title Title of the plot
#' @param p Plot to be added to, if none provided a new plot will be created.
#' @param sections Style used to show sections of the profile.
#' See \code{\link{get_profile}} for description of profile sections.
#'
#' The style options are:
#' \itemize{
#' \item "shade" : shade the background of the different sections different colors
#' \item "lines" : add lines to separate the different sections
#' }
#' @param sec_opacity The opacity of the sections of the profile added to the plot.
#' @return Plotly object.
#' @examples
#' # Plot two profiles
#' profile1 <- get_profile(df, 500, 500, 500)
#' profile2 <- get_profile(df, 200, 300, 500)
#' plot_profiles(list(profile1, profile2), list("profile 1" "profile 2"))
#'
#' # Plot many profiles
#' profiles <- get_profiles(df, 200, 300, 500)
#' plot_profiles(profiles, names(profiles), sections = "shade")
#'
#' # Plot aggregate profiles
#' profiles1 <- get_profiles(dfs, 500, 500, 500)
#' profiles2 <- get_profiles(df, 200, 300, 500)
#' agg_profile1 <- aggregate_profiles(profiles1)
#' agg_profile2 <- aggregate_profiles(profiles2)
#' plot_profiles(list(agg_profile1, agg_profile2), list("aggregated profile 1",
#' "aggregated profile 2"), title = "Aggregate profiles")
#'
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of a profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_profiles <- function(profiles, labels, title = NULL,
p = plot_ly(), sections = "shade", sec_opacity = 0.3) {
profiles <- profiles[!is.null(profiles)]
last <- length(profiles)
for (i in c(1:(last-1))) {
p <- plot_profile(profiles[[i]], labels[[i]],
title = title, color = get_color(i-1), p = p,
sections = NULL)
}
# plot last profile with shading
p <- plot_profile(profiles[[last]], labels[[last]],
title = title, color = get_color(last-1), p = p,
sections = sections, sec_opacity = sec_opacity)
return(p)
}
#' Add section lines and/or shading
#' @description add shading or lines to show the sections of a profile for a plot
#'
#' @param p Plot to add shading to
#' @param scale The number of data points in each third of the profile.
#' See \code{\link{get_profile}} for a description of profile and scale.
#' @param shade Add shading of sections or else add lines separating sections
#' @param ymin Minimum value of the y axis
#' @param ymax Maximum value of the y axis
#' @param opacity The opacity of the shading
#' @return Ploltly object that has section shading or lines
#' @examples
#' profile <- get_profile(df, 300, 200, 500)
#' p <- plot_profile(profile, "the best profile", sections = NULL)
#' add_sections(p, scale = 500)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile and scale.
#' @export
add_sections <- function(p, scale, shade = FALSE, ymin = NULL, ymax = NULL, opacity = 0.3) {
# calculate yman and ymax
if (is.null(ymin) || is.null(ymax)) {
bounds <- min_max_plot(p)
ymin <- if (is.null(ymin)) bounds$min else ymin
ymax <- if (is.null(ymax)) bounds$max else ymax
}
# opacity to 1 if lines
opacity <- if (shade) opacity else 1
fillCol <- function(i) {
colors <- list("tomato", "cornflowerblue", "chartreuse2", "tomato3", "tomato4")
col_code <- col2rgb(colors[[i %% length(colors) + 1]])/255
rgb(col_code[[1]], col_code[[2]], col_code[[3]], opacity)
}
restPre <- c(rep(ymax, scale), rep(NA, scale *2))
active <- c(rep(NA, scale), rep(ymax, scale*0.5), rep(NA, scale*1.5))
rest3min <- c(rep(NA, scale*1.5), rep(ymax, scale*0.5), rep(NA, scale))
restPost <- c(rep(NA, scale*2), rep(ymax, scale))
if (!shade) {
p <- add_lines(p, y = c(ymin:ymax), x = scale, name = "Resting",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(1), width = 3),
hoverinfo = "none")
p <- add_lines(p, y = c(ymin:ymax), x = (scale * 1.5), name = "Active",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(3), width = 3),
hoverinfo = "none")
p <- add_lines(p, y = c(ymin:ymax), x = (scale * 2), name = "Initial Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(2), width = 3),
hoverinfo = "none")
p <- layout(p, yaxis = list(range = c(ymin, ymax)))
} else {
p <- add_lines(p, y = restPre, x = c(1:length(restPre)), name = "Resting",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none", fill = "tozeroy", fillcolor = fillCol(1)) %>%
add_lines(y = active, x = c(1:length(active)), name = "Active",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(3)) %>%
add_lines(y = rest3min, x = c(1:length(rest3min)), name = "Initial Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(2)) %>%
add_lines(y = restPost, x = c(1:length(restPost)), name = "Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(1)) %>%
layout(yaxis = list(range = c(ymin, ymax)))
}
return(p)
}
#' Plot all the activities
#' @description finds and plots all the activities in a energy data frame
#'
#' @param energy_df A dataframe with "energy" format as returned by \code{\link{load_tcx_file}}.
#' @param p A plot to be added to.
#' @param time_threshold The minumum number of seconds duration for an activity.
#' @param rest_threshold The maximum number of resting seconds within and activity.
#' @param type The type of data plotted:
#' \itemize{
#' \item "bpm" : plots the heart rate data in bpm
#' \item "deltaE" : plots the change in energy expenditure
#' \item "totalE" : plots the total energy expenditure
#' }
#' @param real_time Plots data in real time or else uses indexes.
#' @param plot_all Plots the activities and all the data from the data frame.
#' @return A plotly object.
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for details on input format.
#' @export
plot_active <- function(energy_df, p = plot_ly(), time_threshold = 15, rest_threshold = 5, type = "bpm",
real_time = TRUE, plot_all = TRUE) {
# find starting and ending points for all activities in the data frame
start_end_points <- all_start_end_points(energy_df, time_threshold, rest_threshold)
all_points <- start_end_points$points
df <- start_end_points$df
# plot all the data not just activities
if (plot_all) {
if (real_time) index <- df$time
else index <- c(1:nrow(df))
if (type == "bpm") data <- df$bpm
else if (type == "deltaE") data <- df$deltaE
else if (type == "totalE") data <- df$totalE
else stop("type not found")
p <- add_lines(p, y = data, x = index, name = type)
}
# if there are no activities
if(length(all_points) == 0) {
warning("no activities")
return(p)
}
# plot new line for each activity
for(i in c(1:nrow(all_points))) {
points <- all_points[i,]
if (type == "bpm") data <- df[points$start_point:points$end_point,]$bpm
else if (type == "deltaE") data <- df[points$start_point:points$end_point,]$deltaE
else if (type == "totalE")data <- df[points$start_point:points$end_point,]$totalE
else stop("type not found")
if (real_time) index <- df[points$start_point:points$end_point,]$time
else index <- c(points$start_point:points$end_point)
p <- add_trace(p, x = index, y = data,
type = "scatter", mode = "lines",
name = paste("activity ", i))
}
return(p)
}
#' Plots all the features of a profile
#' @description plots:
#' \itemize{
#' \item resting heart rate
#' \item active heart rate
#' \item initial heart rate recovery
#' \item final heart rate recovery
#' \item heart rate recovery every minute post cessation of exercise
#' }
#'
#' @param profile A profile to get features from
#' See \code{\link{get_profile}} for a description of profile.
#' @param buffer_end The number seconds before activity included in profile.
#' @param p Plot to be added to.
#' @param label of the profile being plotted
#' @param join_feat plots all the features as one color and legend group, useful for plotting
#' features of multiple profiles
#' @param col Color of the features when join_feat is true
#' @param annotate_max_hr Add an annotation to the maximum heart rate
#' @return plot with the features of a profile
#' @examples
#' profile1 <- get_profile(df1, 300, 200, 500)
#' plot_features(profile1, 200)
#'
#' profile2 <- get_profile(df2, 300, 200, 500)
#' plot_features(profile2, 200, label = "The best features")
#'
#' p <- plot_features(profile1, 200, join_feat = TRUE)
#' plot_features(profile2, 200, p = p, join_feat = TRUE, col = "orange")
#'
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' @export
plot_features_profile <- function(profile, buffer_end, p = NULL, label = "Features", join_feat = FALSE,
col = get_color(0), annotate_max_hr = TRUE) {
mean = if (is.atomic(profile)) profile else profile$mean
scale = length(mean)/3
# max heart rate annotation
y_val_max <- max_hr(mean)
x_val_max <- match(y_val_max, mean)
# get features as plottable objects
# resting HR
r <- resting_hr(mean)
r <- c(rep(r, scale), rep(NA, scale*2))
# active polynomial
p1 <- as.numeric(predict(profile_coefficients(mean, start=scale, end=scale*1.5, degree=5, model_out = TRUE)))
p1 <- c(rep(NA, (scale - 1)), p1, rep(NA, (scale*1.5)))
# resting polynomial
p2 <- as.numeric(predict(profile_coefficients(mean, start=scale*1.5, end=scale*2, degree=5, model_out = TRUE)))
p2 <- c(rep(NA, (scale*1.5 - 1)), p2, rep(NA, scale))
# slope
s <- as.numeric(predict(slope(mean, model_out = TRUE)))
s <- c(rep(NA, (scale*2 - 1)), s)
# resting indicies
x <- round(scale*1.5/18)
gap <- 42
yvals <- resting_indeces(mean, buffer_end/60, buffer_end)
xvals <- which(mean %in% yvals)
inds <- rep(NA, scale*3)
inds[xvals] <- yvals
# style helper functions
line_style <- function(i) {
# colors <- list("tomato", "tomato1", "tomato2", "tomato3", "tomato4")
color <- if (join_feat) col else get_color(i)
list(
color = color,
dash = "line",
width = 2
)
}
legend_group <- if (join_feat) paste(label, "feat") else ""
show_legend <- if (join_feat) FALSE else TRUE
first_name <- if (join_feat) label else "Resting Heart Rate"
# plot all of the features
if (is.null(p)) {
title <- "Heart Rate Profile Features"
title <- if (label != "Features") paste(label, title) else title
p <- plot_ly(x = c(1:length(mean)), y = r, name = first_name, text = "",
mode = "lines", legendgroup = legend_group, showlegend = TRUE,
line = line_style(0), type = "scatter") %>%
layout(title = title)
} else {
p <- add_lines(p, x = c(1:length(r)), y = r, name = first_name, text = "",
legendgroup = legend_group, showlegend = TRUE,
line = line_style(0))
}
# add features
p <- add_lines(p, x = c(1:length(p1)), y = p1, name = "Active", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(1)) %>%
add_lines(x = c(1:length(p2)), y = p2, name = "3 min Resting", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(2)) %>%
add_lines(x = c(1:length(s)), y = s, name = "Recovery", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(3)) %>%
add_markers(x = xvals, y = yvals, name = "Resting Indices",
text = "", inherit = FALSE,
marker = list(
color = if (join_feat) col else get_color(4),
line = list(color = 'rgba(0,0,0,0)', width = 0)),
legendgroup = legend_group, showlegend = show_legend) %>%
add_markers(x = x_val_max, y = y_val_max, name = "Maximum Heart Rate",
text = "", inherit = FALSE,
marker = list(
color = if (join_feat) col else get_color(5),
line = list(color = 'rgba(0,0,0,0)', width = 0)),
legendgroup = legend_group, showlegend = show_legend)
# add layout options
p <- layout(p, hovermode = "compare",
yaxis = list(title = "BPM"),
xaxis = list(title = "Sections"))
return(p)
}
#' Plot features of multiple profiles
#' @description plot the features of all the profiles in a list on the same plot,
#' to easily compare between features from different profiles
#'
#' @param profiles list of profiles or aggregate profiles to be plotted
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param buffer_end The number seconds before activity included in profile.
#' @return Plotly plot
#' @examples
#' profiles <- get_profiles(dfs, 200, 300, 500)
#' plot_features_profiles(profiles, names(profiles), 300)
#'
#' agg_profile1 <- aggregate_profiles(select(profiles, "1"))
#' agg_profile2 <- aggregate_profiles(select(profiles, "2"))
#' plot_features_profiles(list(agg_profile1, agg_profile2), list("1", "2"), 300)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_features_profiles <- function(profiles, labels, buffer_end) {
p <- plot_ly()
for (i in c(1:length(profiles))){
p <- plot_features_profile(profiles[[i]], buffer_end, p = p, label = labels[[i]],
join_feat = TRUE, col = get_color(i), annotate_max_hr = FALSE)
}
p <- layout(p, title="Features of All Profiles",
yaxis = list(title = "BPM"),
xaxis = list(title = "Sections"))
return(p)
}
#' Plot multiple profiles in 3d based on energy
#' @description plot multiple profiles in 3d against energy
#'
#' @inheritParams get_energy_profiles
#' @param plot_energy Plot energy curve instead of heart rate profile
#' @return Plotly plot
#' @examples plot_3d_profiles(energy_dfs, buffer_start, buffer_end, scale)
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for an explanation of energy dataframes
#' @export
plot_3d_profiles <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15, plot_energy = FALSE) {
# step 1, get the profiles in a usable form
# get profiles
energy_profiles <- get_energy_profiles(energy_dfs, buffer_start, buffer_end,
scale, time_threshold = time_threshold)
# sort
e_order <- order(unlist(lapply(energy_profiles, function(x) x$energy)))
activities <- energy_profiles[e_order]
# step 2, plot
p <- plot_ly()
line_type <- list(width = 6) #, color = ~BPM, colorscale = 'Viridis')
for (i in c(1:length(activities))) {
activity <- activities[[i]]
name <- names(activities)[[i]]
if (plot_energy) {
data <- data.frame(
Index = c(1:length(activity$totalE)),
EnergyRank = activity$energy,
Energy = activity$totalE
)
p <- add_trace(p, data = data, y = ~EnergyRank, x = ~Index, z = ~Energy,
line = line_type, name = paste(name, activity$energy),
type = "scatter3d", mode = "lines")
} else {
data <- data.frame(
Index = c(1:length(activity$profile)),
EnergyRank = activity$energy,
BPM = activity$profile
)
p <- add_trace(p, data = data, y = ~EnergyRank, x = ~Index, z = ~BPM,
line = line_type, name = name,
type = "scatter3d", mode = "lines")
}
}
return(p)
}
#' Plot multiple profiles in 3d landscape based on energy
#' @description plot multiple profiles in 3d landscape by interpolating heart rate data points
#'
#' @inheritParams get_energy_profiles
#' @return Plotly plot
#' @examples plot_3d_landscape(energy_dfs, buffer_start, buffer_end, scale)
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for an explanation of energy dataframes.
#' @export
plot_3d_landscape <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15, plot_energy = FALSE) {
# step 1, get the profiles in a usable form
# get profiles
energy_profiles <- get_energy_profiles(energy_dfs, buffer_start, buffer_end,
scale, time_threshold = time_threshold)
# sort
e_order <- order(unlist(lapply(energy_profiles, function(x) x$energy)))
energies_ord <- energy_profiles[e_order]
# step 2, interpolate
es <- unlist(lapply(energies_ord, function(x) x$energy))
matrix_indices = seq(from=round(range(es)[1]), to=round(range(es)[2]), by=1)
profile_matrix <- matrix(NA, length(matrix_indices), scale*3)
rownames(profile_matrix) <- matrix_indices
lapply(energies_ord, function(X){
tmpindex <- which(rownames(profile_matrix)==round(X$energy))
profile_matrix[tmpindex,] <<- X$profile
return("Yay")
})
interpolated_profile_matrix <- apply(profile_matrix, 2, function(Y){ return(approx(Y, n=length(Y))$y)})
rownames(interpolated_profile_matrix) <- rownames(profile_matrix)
# step 3, plot
p <- plot_ly(y = as.numeric(rownames(interpolated_profile_matrix)),
z = ~interpolated_profile_matrix) %>% add_surface()
return(p)
}
# plot_3d_profiles helper functions
#' Get profiles from energy dataframes with added information ready for plotting
#' with plot_3d_profiles
#' @description prepare data to be plotted by plot_3d_profiles
#'
#' @param energy_dfs list of profiles or aggregate profiles to be plotted
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @inheritParams get_profile
#' @return List of dataframes with profiles and other useful information
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
get_energy_profiles <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15) {
activities <- lapply(energy_dfs, function(x) {
act <- get_activities(x, buffer_start = buffer_start, buffer_end = buffer_end,
scale = scale, time_threshold = time_threshold,
all_activities = FALSE)
})
activities <- unlist(activities, recursive = FALSE)
profiles <- lapply(activities, function(x) {
result <- list()
points <- list("start_point" = buffer_start + 1,
"end_point" = nrow(x) - buffer_end)
range <- c((points$start_point-1):(points$end_point+1))
result$profile <- get_profile(x, buffer_start = buffer_start,
buffer_end = buffer_end, scale = scale,
points = points)
# calc energy area
time <- as.numeric(x$time[range])
energy <- x$totalE[range]
quotient <- mapply('+', energy[1:length(energy)-1],
energy[2:length(energy)])/2
product <- quotient * diff(time)
time_diff <- range(x$time[range])
# result$energy <- sum(product)
# result$energy <- mean(energy)
result$energy <- max(energy)
result$totalE <- x$totalE
result$bpm <- x$bpm
result$range <- range
result$time <- x$time
return(result)
})
return(profiles)
}
VCCRI/CardiacProfileR documentation built on May 25, 2019, 2:22 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# plotting.R
# functions that can plot profiles
####################
# helper functions #
####################
#' Given an index returns a plotly color
#' @description returns a plotly color to be used in plotting functions
#'
#' @param i Index of color
#' @return Color to be used in plotting function
#' @family plotting functions
get_color <- function(i) {
colors <- list(
'#1f77b4', # muted blue
'#ff7f0e', # safety orange
'#2ca02c', # cooked asparagus green
'#d62728', # brick red
'#9467bd', # muted purple
'#8c564b', # chestnut brown
'#e377c2', # raspberry yogurt pink
'#7f7f7f', # middle gray
'#bcbd22', # curry yellow-green
'#17becf' # blue-teal
)
nc = length(colors) # number of colors
return(colors[[i %% nc + 1]])
}
#' Get ymin ymax from plot
#' @description returns the minimum and maximum y values from a plotly function
#'
#' @param p Plotly plot
#' @return list with:
#' \itemize{
#' \item min : minimum y value
#' \item max : maximum y value
#' }
#' @family plotting functions
min_max_plot <- function(p) {
yvals <- unlist(lapply(p$x$attrs, function(x) {
if ("y" %in% names(x)) return(x$y)
if ("ymin" %in% names(x)) return(c(x$ymin, x$ymax))
}))
yvals <- yvals[!is.na(yvals)]
bounds <- list("min" = floor(min(yvals))-1, "max" = ceiling(max(yvals))+1)
return(bounds)
}
#####################
# plotting function #
#####################
#' Plot a single profile
#' @description plot single profiles individually or add to existing plot
#'
#' @param profile Profile or aggregate profile to be plotted.
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param label Label of the profile being plotted
#' @param title Title of the plot
#' @param color Color of the profile added to the plot.
#' @param p Plot to be added to, if none provided a new plot will be created.
#' @param sections Style used to plot sections of the profile.
#' See \code{\link{get_profile}} for description of profile sections.
#'
#' The style options are:
#' \itemize{
#' \item "shade" : shade the background of the different sections different colors
#' \item "lines" : add lines to separate the different sections
#' }
#' @param sec_opacity The opacity of the sections of the profile added to the plot.
#' @param ribbon_opacity The opacity of standard deviation ribbon when plotting aggregate profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @return Plotly object.
#' @examples
#' profile1 <- get_profile(df, 500, 500, 500)
#' p <- plot_profile(profile1, "profile 1", title = "Great Profiles")
#' profile2 <- get_profile(df, 200, 300, 500)
#' plot_profile(profile2, "profile 2", p = p)
#'
#' profiles <- get_profiles(dfs, 500, 500, 500)
#' agg_profile <- aggregate_profiles(profiles)
#' plot_profile(agg_profile, "aggregated profile", ribbon_opacity = 0.2)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_profile <- function(profile, label, title = NULL, color = get_color(0),
p = plot_ly(), sections = "shade",
sec_opacity = 0.3, ribbon_opactiy = 0.4) {
# check if it's an aggregate profile or not
if (Reduce("&", c("mean", "sd") %in% names(profile))) {
is_aggregate = TRUE
} else {
is_aggregate = FALSE
}
# plot data
profile <- data.frame(profile)
xval = c(1:nrow(profile))
if (is_aggregate) {
# plot mean line
p <- add_lines(p, x = xval, y = profile$mean,
text = paste('sd: ', profile$sd),
line = list(color = color),
legendgroup = label, name = label)
# add sd ribbon
p <- add_ribbons(p, x = xval,
ymin = profile$mean - profile$sd,
ymax = profile$mean + profile$sd,
line = list(color = color),
opacity = ribbon_opactiy, hoverinfo = "none",
legendgroup = label, name = label,
inherit = TRUE, showlegend = FALSE)
} else {
# plot individual profile
p <- add_lines(p, x = xval, y = profile$profile, name = label)
}
p <- layout(p,
xaxis = list(title = "Sections", showticklabels = FALSE),
yaxis = list(title = "BPM"),
title = title)
# determine type of sectioning
sections <- if (is.null(sections)) 'none' else sections
section_option <- list("lines" = FALSE, "shade" = TRUE)[[sections]]
# add sections (either as lines or shading)
if (!is.null(section_option)) {
p <- add_sections(p, shade = section_option, opacity = sec_opacity, scale = nrow(profile)/3)
}
return(p)
}
#' Plot a many individual profiles
#' @description plot many profiles or aggregate profiles on one plot
#'
#' @param profiles A list of profiles or aggregate profiles to be plotted.
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param labels A list of labels of the profiles being plotted.
#' @param title Title of the plot
#' @param p Plot to be added to, if none provided a new plot will be created.
#' @param sections Style used to show sections of the profile.
#' See \code{\link{get_profile}} for description of profile sections.
#'
#' The style options are:
#' \itemize{
#' \item "shade" : shade the background of the different sections different colors
#' \item "lines" : add lines to separate the different sections
#' }
#' @param sec_opacity The opacity of the sections of the profile added to the plot.
#' @return Plotly object.
#' @examples
#' # Plot two profiles
#' profile1 <- get_profile(df, 500, 500, 500)
#' profile2 <- get_profile(df, 200, 300, 500)
#' plot_profiles(list(profile1, profile2), list("profile 1" "profile 2"))
#'
#' # Plot many profiles
#' profiles <- get_profiles(df, 200, 300, 500)
#' plot_profiles(profiles, names(profiles), sections = "shade")
#'
#' # Plot aggregate profiles
#' profiles1 <- get_profiles(dfs, 500, 500, 500)
#' profiles2 <- get_profiles(df, 200, 300, 500)
#' agg_profile1 <- aggregate_profiles(profiles1)
#' agg_profile2 <- aggregate_profiles(profiles2)
#' plot_profiles(list(agg_profile1, agg_profile2), list("aggregated profile 1",
#' "aggregated profile 2"), title = "Aggregate profiles")
#'
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of a profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_profiles <- function(profiles, labels, title = NULL,
p = plot_ly(), sections = "shade", sec_opacity = 0.3) {
profiles <- profiles[!is.null(profiles)]
last <- length(profiles)
for (i in c(1:(last-1))) {
p <- plot_profile(profiles[[i]], labels[[i]],
title = title, color = get_color(i-1), p = p,
sections = NULL)
}
# plot last profile with shading
p <- plot_profile(profiles[[last]], labels[[last]],
title = title, color = get_color(last-1), p = p,
sections = sections, sec_opacity = sec_opacity)
return(p)
}
#' Add section lines and/or shading
#' @description add shading or lines to show the sections of a profile for a plot
#'
#' @param p Plot to add shading to
#' @param scale The number of data points in each third of the profile.
#' See \code{\link{get_profile}} for a description of profile and scale.
#' @param shade Add shading of sections or else add lines separating sections
#' @param ymin Minimum value of the y axis
#' @param ymax Maximum value of the y axis
#' @param opacity The opacity of the shading
#' @return Ploltly object that has section shading or lines
#' @examples
#' profile <- get_profile(df, 300, 200, 500)
#' p <- plot_profile(profile, "the best profile", sections = NULL)
#' add_sections(p, scale = 500)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile and scale.
#' @export
add_sections <- function(p, scale, shade = FALSE, ymin = NULL, ymax = NULL, opacity = 0.3) {
# calculate yman and ymax
if (is.null(ymin) || is.null(ymax)) {
bounds <- min_max_plot(p)
ymin <- if (is.null(ymin)) bounds$min else ymin
ymax <- if (is.null(ymax)) bounds$max else ymax
}
# opacity to 1 if lines
opacity <- if (shade) opacity else 1
fillCol <- function(i) {
colors <- list("tomato", "cornflowerblue", "chartreuse2", "tomato3", "tomato4")
col_code <- col2rgb(colors[[i %% length(colors) + 1]])/255
rgb(col_code[[1]], col_code[[2]], col_code[[3]], opacity)
}
restPre <- c(rep(ymax, scale), rep(NA, scale *2))
active <- c(rep(NA, scale), rep(ymax, scale*0.5), rep(NA, scale*1.5))
rest3min <- c(rep(NA, scale*1.5), rep(ymax, scale*0.5), rep(NA, scale))
restPost <- c(rep(NA, scale*2), rep(ymax, scale))
if (!shade) {
p <- add_lines(p, y = c(ymin:ymax), x = scale, name = "Resting",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(1), width = 3),
hoverinfo = "none")
p <- add_lines(p, y = c(ymin:ymax), x = (scale * 1.5), name = "Active",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(3), width = 3),
hoverinfo = "none")
p <- add_lines(p, y = c(ymin:ymax), x = (scale * 2), name = "Initial Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = fillCol(2), width = 3),
hoverinfo = "none")
p <- layout(p, yaxis = list(range = c(ymin, ymax)))
} else {
p <- add_lines(p, y = restPre, x = c(1:length(restPre)), name = "Resting",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none", fill = "tozeroy", fillcolor = fillCol(1)) %>%
add_lines(y = active, x = c(1:length(active)), name = "Active",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(3)) %>%
add_lines(y = rest3min, x = c(1:length(rest3min)), name = "Initial Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(2)) %>%
add_lines(y = restPost, x = c(1:length(restPost)), name = "Recovery",
legendgroup = "Sections", text = "",
mode = "none", line = list(color = rgb(0, 0, 0, 0)),
hoverinfo = "none",
fill = "tozeroy", fillcolor = fillCol(1)) %>%
layout(yaxis = list(range = c(ymin, ymax)))
}
return(p)
}
#' Plot all the activities
#' @description finds and plots all the activities in a energy data frame
#'
#' @param energy_df A dataframe with "energy" format as returned by \code{\link{load_tcx_file}}.
#' @param p A plot to be added to.
#' @param time_threshold The minumum number of seconds duration for an activity.
#' @param rest_threshold The maximum number of resting seconds within and activity.
#' @param type The type of data plotted:
#' \itemize{
#' \item "bpm" : plots the heart rate data in bpm
#' \item "deltaE" : plots the change in energy expenditure
#' \item "totalE" : plots the total energy expenditure
#' }
#' @param real_time Plots data in real time or else uses indexes.
#' @param plot_all Plots the activities and all the data from the data frame.
#' @return A plotly object.
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for details on input format.
#' @export
plot_active <- function(energy_df, p = plot_ly(), time_threshold = 15, rest_threshold = 5, type = "bpm",
real_time = TRUE, plot_all = TRUE) {
# find starting and ending points for all activities in the data frame
start_end_points <- all_start_end_points(energy_df, time_threshold, rest_threshold)
all_points <- start_end_points$points
df <- start_end_points$df
# plot all the data not just activities
if (plot_all) {
if (real_time) index <- df$time
else index <- c(1:nrow(df))
if (type == "bpm") data <- df$bpm
else if (type == "deltaE") data <- df$deltaE
else if (type == "totalE") data <- df$totalE
else stop("type not found")
p <- add_lines(p, y = data, x = index, name = type)
}
# if there are no activities
if(length(all_points) == 0) {
warning("no activities")
return(p)
}
# plot new line for each activity
for(i in c(1:nrow(all_points))) {
points <- all_points[i,]
if (type == "bpm") data <- df[points$start_point:points$end_point,]$bpm
else if (type == "deltaE") data <- df[points$start_point:points$end_point,]$deltaE
else if (type == "totalE")data <- df[points$start_point:points$end_point,]$totalE
else stop("type not found")
if (real_time) index <- df[points$start_point:points$end_point,]$time
else index <- c(points$start_point:points$end_point)
p <- add_trace(p, x = index, y = data,
type = "scatter", mode = "lines",
name = paste("activity ", i))
}
return(p)
}
#' Plots all the features of a profile
#' @description plots:
#' \itemize{
#' \item resting heart rate
#' \item active heart rate
#' \item initial heart rate recovery
#' \item final heart rate recovery
#' \item heart rate recovery every minute post cessation of exercise
#' }
#'
#' @param profile A profile to get features from
#' See \code{\link{get_profile}} for a description of profile.
#' @param buffer_end The number seconds before activity included in profile.
#' @param p Plot to be added to.
#' @param label of the profile being plotted
#' @param join_feat plots all the features as one color and legend group, useful for plotting
#' features of multiple profiles
#' @param col Color of the features when join_feat is true
#' @param annotate_max_hr Add an annotation to the maximum heart rate
#' @return plot with the features of a profile
#' @examples
#' profile1 <- get_profile(df1, 300, 200, 500)
#' plot_features(profile1, 200)
#'
#' profile2 <- get_profile(df2, 300, 200, 500)
#' plot_features(profile2, 200, label = "The best features")
#'
#' p <- plot_features(profile1, 200, join_feat = TRUE)
#' plot_features(profile2, 200, p = p, join_feat = TRUE, col = "orange")
#'
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' @export
plot_features_profile <- function(profile, buffer_end, p = NULL, label = "Features", join_feat = FALSE,
col = get_color(0), annotate_max_hr = TRUE) {
mean = if (is.atomic(profile)) profile else profile$mean
scale = length(mean)/3
# max heart rate annotation
y_val_max <- max_hr(mean)
x_val_max <- match(y_val_max, mean)
# get features as plottable objects
# resting HR
r <- resting_hr(mean)
r <- c(rep(r, scale), rep(NA, scale*2))
# active polynomial
p1 <- as.numeric(predict(profile_coefficients(mean, start=scale, end=scale*1.5, degree=5, model_out = TRUE)))
p1 <- c(rep(NA, (scale - 1)), p1, rep(NA, (scale*1.5)))
# resting polynomial
p2 <- as.numeric(predict(profile_coefficients(mean, start=scale*1.5, end=scale*2, degree=5, model_out = TRUE)))
p2 <- c(rep(NA, (scale*1.5 - 1)), p2, rep(NA, scale))
# slope
s <- as.numeric(predict(slope(mean, model_out = TRUE)))
s <- c(rep(NA, (scale*2 - 1)), s)
# resting indicies
x <- round(scale*1.5/18)
gap <- 42
yvals <- resting_indeces(mean, buffer_end/60, buffer_end)
xvals <- which(mean %in% yvals)
inds <- rep(NA, scale*3)
inds[xvals] <- yvals
# style helper functions
line_style <- function(i) {
# colors <- list("tomato", "tomato1", "tomato2", "tomato3", "tomato4")
color <- if (join_feat) col else get_color(i)
list(
color = color,
dash = "line",
width = 2
)
}
legend_group <- if (join_feat) paste(label, "feat") else ""
show_legend <- if (join_feat) FALSE else TRUE
first_name <- if (join_feat) label else "Resting Heart Rate"
# plot all of the features
if (is.null(p)) {
title <- "Heart Rate Profile Features"
title <- if (label != "Features") paste(label, title) else title
p <- plot_ly(x = c(1:length(mean)), y = r, name = first_name, text = "",
mode = "lines", legendgroup = legend_group, showlegend = TRUE,
line = line_style(0), type = "scatter") %>%
layout(title = title)
} else {
p <- add_lines(p, x = c(1:length(r)), y = r, name = first_name, text = "",
legendgroup = legend_group, showlegend = TRUE,
line = line_style(0))
}
# add features
p <- add_lines(p, x = c(1:length(p1)), y = p1, name = "Active", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(1)) %>%
add_lines(x = c(1:length(p2)), y = p2, name = "3 min Resting", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(2)) %>%
add_lines(x = c(1:length(s)), y = s, name = "Recovery", text = "",
legendgroup = legend_group, showlegend = show_legend,
line = line_style(3)) %>%
add_markers(x = xvals, y = yvals, name = "Resting Indices",
text = "", inherit = FALSE,
marker = list(
color = if (join_feat) col else get_color(4),
line = list(color = 'rgba(0,0,0,0)', width = 0)),
legendgroup = legend_group, showlegend = show_legend) %>%
add_markers(x = x_val_max, y = y_val_max, name = "Maximum Heart Rate",
text = "", inherit = FALSE,
marker = list(
color = if (join_feat) col else get_color(5),
line = list(color = 'rgba(0,0,0,0)', width = 0)),
legendgroup = legend_group, showlegend = show_legend)
# add layout options
p <- layout(p, hovermode = "compare",
yaxis = list(title = "BPM"),
xaxis = list(title = "Sections"))
return(p)
}
#' Plot features of multiple profiles
#' @description plot the features of all the profiles in a list on the same plot,
#' to easily compare between features from different profiles
#'
#' @param profiles list of profiles or aggregate profiles to be plotted
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @param buffer_end The number seconds before activity included in profile.
#' @return Plotly plot
#' @examples
#' profiles <- get_profiles(dfs, 200, 300, 500)
#' plot_features_profiles(profiles, names(profiles), 300)
#'
#' agg_profile1 <- aggregate_profiles(select(profiles, "1"))
#' agg_profile2 <- aggregate_profiles(select(profiles, "2"))
#' plot_features_profiles(list(agg_profile1, agg_profile2), list("1", "2"), 300)
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @export
plot_features_profiles <- function(profiles, labels, buffer_end) {
p <- plot_ly()
for (i in c(1:length(profiles))){
p <- plot_features_profile(profiles[[i]], buffer_end, p = p, label = labels[[i]],
join_feat = TRUE, col = get_color(i), annotate_max_hr = FALSE)
}
p <- layout(p, title="Features of All Profiles",
yaxis = list(title = "BPM"),
xaxis = list(title = "Sections"))
return(p)
}
#' Plot multiple profiles in 3d based on energy
#' @description plot multiple profiles in 3d against energy
#'
#' @inheritParams get_energy_profiles
#' @param plot_energy Plot energy curve instead of heart rate profile
#' @return Plotly plot
#' @examples plot_3d_profiles(energy_dfs, buffer_start, buffer_end, scale)
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for an explanation of energy dataframes
#' @export
plot_3d_profiles <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15, plot_energy = FALSE) {
# step 1, get the profiles in a usable form
# get profiles
energy_profiles <- get_energy_profiles(energy_dfs, buffer_start, buffer_end,
scale, time_threshold = time_threshold)
# sort
e_order <- order(unlist(lapply(energy_profiles, function(x) x$energy)))
activities <- energy_profiles[e_order]
# step 2, plot
p <- plot_ly()
line_type <- list(width = 6) #, color = ~BPM, colorscale = 'Viridis')
for (i in c(1:length(activities))) {
activity <- activities[[i]]
name <- names(activities)[[i]]
if (plot_energy) {
data <- data.frame(
Index = c(1:length(activity$totalE)),
EnergyRank = activity$energy,
Energy = activity$totalE
)
p <- add_trace(p, data = data, y = ~EnergyRank, x = ~Index, z = ~Energy,
line = line_type, name = paste(name, activity$energy),
type = "scatter3d", mode = "lines")
} else {
data <- data.frame(
Index = c(1:length(activity$profile)),
EnergyRank = activity$energy,
BPM = activity$profile
)
p <- add_trace(p, data = data, y = ~EnergyRank, x = ~Index, z = ~BPM,
line = line_type, name = name,
type = "scatter3d", mode = "lines")
}
}
return(p)
}
#' Plot multiple profiles in 3d landscape based on energy
#' @description plot multiple profiles in 3d landscape by interpolating heart rate data points
#'
#' @inheritParams get_energy_profiles
#' @return Plotly plot
#' @examples plot_3d_landscape(energy_dfs, buffer_start, buffer_end, scale)
#' @family plotting functions
#' @seealso \code{\link{load_tcx_file}} for an explanation of energy dataframes.
#' @export
plot_3d_landscape <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15, plot_energy = FALSE) {
# step 1, get the profiles in a usable form
# get profiles
energy_profiles <- get_energy_profiles(energy_dfs, buffer_start, buffer_end,
scale, time_threshold = time_threshold)
# sort
e_order <- order(unlist(lapply(energy_profiles, function(x) x$energy)))
energies_ord <- energy_profiles[e_order]
# step 2, interpolate
es <- unlist(lapply(energies_ord, function(x) x$energy))
matrix_indices = seq(from=round(range(es)[1]), to=round(range(es)[2]), by=1)
profile_matrix <- matrix(NA, length(matrix_indices), scale*3)
rownames(profile_matrix) <- matrix_indices
lapply(energies_ord, function(X){
tmpindex <- which(rownames(profile_matrix)==round(X$energy))
profile_matrix[tmpindex,] <<- X$profile
return("Yay")
})
interpolated_profile_matrix <- apply(profile_matrix, 2, function(Y){ return(approx(Y, n=length(Y))$y)})
rownames(interpolated_profile_matrix) <- rownames(profile_matrix)
# step 3, plot
p <- plot_ly(y = as.numeric(rownames(interpolated_profile_matrix)),
z = ~interpolated_profile_matrix) %>% add_surface()
return(p)
}
# plot_3d_profiles helper functions
#' Get profiles from energy dataframes with added information ready for plotting
#' with plot_3d_profiles
#' @description prepare data to be plotted by plot_3d_profiles
#'
#' @param energy_dfs list of profiles or aggregate profiles to be plotted
#' See \code{\link{get_profile}} for a description of profile.
#' See \code{\link{aggregate_profiles}} for a description of an aggregate profile.
#' @inheritParams get_profile
#' @return List of dataframes with profiles and other useful information
#' @family plotting functions
#' @seealso \code{\link{get_profile}} for a description of profile.
#' \code{\link{aggregate_profiles}} for a description of an aggregate profile.
get_energy_profiles <- function(energy_dfs, buffer_start, buffer_end, scale,
time_threshold = 15) {
activities <- lapply(energy_dfs, function(x) {
act <- get_activities(x, buffer_start = buffer_start, buffer_end = buffer_end,
scale = scale, time_threshold = time_threshold,
all_activities = FALSE)
})
activities <- unlist(activities, recursive = FALSE)
profiles <- lapply(activities, function(x) {
result <- list()
points <- list("start_point" = buffer_start + 1,
"end_point" = nrow(x) - buffer_end)
range <- c((points$start_point-1):(points$end_point+1))
result$profile <- get_profile(x, buffer_start = buffer_start,
buffer_end = buffer_end, scale = scale,
points = points)
# calc energy area
time <- as.numeric(x$time[range])
energy <- x$totalE[range]
quotient <- mapply('+', energy[1:length(energy)-1],
energy[2:length(energy)])/2
product <- quotient * diff(time)
time_diff <- range(x$time[range])
# result$energy <- sum(product)
# result$energy <- mean(energy)
result$energy <- max(energy)
result$totalE <- x$totalE
result$bpm <- x$bpm
result$range <- range
result$time <- x$time
return(result)
})
return(profiles)
}
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.