,
Matthew Fidler [ctb] ,
Novartis Pharma AG [cph, fnd]| Title: | Exploratory Graphics for Pharmacometrics |
|---|---|
| Description: | Supports a structured approach for exploring PKPD data <https://opensource.nibr.com/xgx/>. It also contains helper functions for enabling the modeler to follow best R practices (by appending the program name, figure name location, and draft status to each plot). In addition, it enables the modeler to follow best graphical practices (by providing a theme that reduces chart ink, and by providing time-scale, log-scale, and reverse-log-transform-scale functions for more readable axes). Finally, it provides some data checking and summarizing functions for rapidly exploring pharmacokinetics and pharmacodynamics (PKPD) datasets. |
| Authors: | Andrew Stein [aut, cre],
Alison Margolskee [aut],
Fariba Khanshan [aut],
Konstantin Krismer [aut] ,
Matthew Fidler [ctb] ,
Novartis Pharma AG [cph, fnd] |
| Maintainer: | Andrew Stein <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 1.1.4 |
| Built: | 2024-11-05 05:30:21 UTC |
| Source: | https://github.com/novartis/xgxr |
Case 1 PKPD Data Set
case1_pkpdcase1_pkpd
A data frame with the following 21 columns:
| column 1: | ID |
integer; unique subject ID |
| column 2: | TIME |
numeric; time relative to first drug administration |
| column 3: | NOMTIME |
numeric; nominal time |
| column 4: | TIMEUNIT |
factor; unit of TIME |
| column 5: | AMT |
integer; dosing amount (for dosing events) in mg |
| column 6: | LIDV |
numeric; observation on a linear scale (observation type determined by CMT), units determined by EVENTU column |
| column 7: | CMT |
integer; compartment number (determines observation type): |
| CMT 1 = Dosing event | ||
| CMT 2 = PK concentration | ||
| CMT 3 = Continuous response data | ||
| CMT 4 = Count response data | ||
| CMT 5 = Ordinal response data | ||
| CMT 6 = Binary response data | ||
| column 8: | NAME |
factor; description of event |
| column 9: | EVENTU |
factor; unit for observation |
| column 10: | CENS |
integer; censored values (0 = not censored, 1 = censored) |
| column 11: | EVID |
integer; event ID (0 = observation, 1 = dosing event) |
| column 12: | WEIGHTB |
numeric; baseline body weight (kg) |
| column 13: | eff0 |
numeric; efficacy |
| column 14: | TRTACT |
factor; treatment group label |
| column 15: | DOSE |
integer; Dose in mg |
| column 16: | PROFDAY |
integer; day of profile |
| column 17: | PROFTIME |
numeric; time within PROFDAY |
| column 18: | CYCLE |
integer; count of drug administrations received |
| column 19: | PART |
integer; part of study |
| column 20: | STUDY |
integer; study |
| column 21: | IPRED |
numeric; individual prediction |
edit_rmd_template_str returns a path to the altered Rmd template
edit_rmd_template_str( rmd_str = NULL, mapping = NULL, rmd_output_path = NULL, data_path = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, dose_cmt = NULL, steady_state_day = NULL, time_between_doses = NULL, author_name = NULL, add_datetime = TRUE, show_explanation = TRUE )edit_rmd_template_str( rmd_str = NULL, mapping = NULL, rmd_output_path = NULL, data_path = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, dose_cmt = NULL, steady_state_day = NULL, time_between_doses = NULL, author_name = NULL, add_datetime = TRUE, show_explanation = TRUE )
rmd_str |
A character string containing the Rmd template raw characters |
mapping |
A list of column name mappings from the original (template) dataset column names to the corresponding columns in the new dataset |
rmd_output_path |
A custom output path for the generated Rmd file (This is typically left as 'NULL' in order to maintain the hierarchical directory structure of 'xgx_autoexplore_output')) |
data_path |
Path (as a string) to the dataset that is to be analyzed |
multiple_dosing |
if FALSE use single ascending dose template, if TRUE use multiple |
pk_cmt |
An integer denoting the "compartment" containing the PK data. The "CMT" column will typically have these integers, where each row may contain either PK or PD data, potentially of different types (continuous, ordinal, etc.) |
pd_cmt |
An integer denoting the "compartment" containing the PD data, of the desired type (continuous, ordinal, etc.). The "CMT" column will typically have these integers, where each row may contain either PK or PD data |
dose_cmt |
CMT associated with dosing event |
steady_state_day |
For multiple ascending dose, what day is steady state rich profile? |
time_between_doses |
time interval between doses |
author_name |
The name of the author to be displayed on the template |
add_datetime |
Boolean indicating additon of a date stamp to the beginnning of the Rmd file |
show_explanation |
Boolean indicating if the additional explanations (text in between figures) are needed for the user. |
A string of the new R markdown template
get_rmd_name returns a name for an Rmd template, based on the desired PKPD parameters
get_rmd_name( rmd_template_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL )get_rmd_name( rmd_template_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL )
rmd_template_name |
A custom output name for the generated Rmd file |
multiple_dosing |
if FALSE use single ascending dose template, if TRUE use multiple |
pk_cmt |
An integer denoting the "compartment" containing the PK data. The "CMT" column will typically have these integers, where each row may contain either PK or PD data, potentially of different types (continuous, ordinal, etc.) |
pd_cmt |
An integer denoting the "compartment" containing the PD data, of the desired type (continuous, ordinal, etc.). The "CMT" column will typically have these integers, where each row may contain either PK or PD data |
pd_data_type |
The type of PD data - acceptable values exist in the following list: ["binary","continuous","count","ordinal","real_example","receptor_occupancy","time_to_event"] |
a string for the Rmd template name
get_rmd_str returns a Rmd template string, based on the desired PKPD parametersget_rmd_str returns a Rmd template string, based on the desired PKPD parameters
get_rmd_str( rmd_template_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL )get_rmd_str( rmd_template_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL )
rmd_template_name |
A custom output name for the generated Rmd file |
multiple_dosing |
if FALSE use single ascending dose template, if TRUE use multiple |
pk_cmt |
An integer denoting the "compartment" containing the PK data. The "CMT" column will typically have these integers, where each row may contain either PK or PD data, potentially of different types (continuous, ordinal, etc.) |
pd_cmt |
An integer denoting the "compartment" containing the PD data, of the desired type (continuous, ordinal, etc.). The "CMT" column will typically have these integers, where each row may contain either PK or PD data |
pd_data_type |
The type of PD data - acceptable values exist in the following list: ["binary","continuous","count","ordinal","real_example","receptor_occupancy","time_to_event"] |
a string for the Rmd template name
Model generated PK and PD data to mimic an orally administered small molecule with various endpoints from continuous to ordinal response and count data. Simulated multiple dose administration ranging from 100 mg to 1600 mg, once per day.
madmad
A data frame with the following 19 columns:
| column 1: | ID |
numeric; unique subject ID |
| column 2: | TIME |
numeric; time relative to first drug administration |
| column 3: | NOMTIME |
numeric; nominal time |
| column 4: | TIMEUNIT |
character; unit of TIME |
| column 5: | AMT |
numeric; dosing amount (for dosing events) in mg |
| column 6: | LIDV |
numeric; observation on a linear scale (observation type determined by CMT), units determined by EVENTU column |
| column 7: | MDV |
numeric; missing dependent variable |
| column 8: | CMT |
integer; compartment number (determines observation type): |
| CMT 1 = Dosing event | ||
| CMT 2 = PK concentration | ||
| CMT 3 = Continuous response data | ||
| CMT 4 = Count response data | ||
| CMT 5 = Ordinal response data | ||
| CMT 6 = Binary response data | ||
| column 9: | NAME |
character; description of event |
| column 10: | EVENTU |
character; unit for observation |
| column 11: | CENS |
integer; censored values (0 = not censored, 1 = censored) |
| column 12: | EVID |
integer; event ID (0 = observation, 1 = dosing event) |
| column 13: | WEIGHTB |
numeric; baseline body weight (kg) |
| column 14: | SEX |
character; sex |
| column 15: | TRTACT |
factor; treatment group label |
| column 16: | DOSE |
numeric; randomized dose in mg |
| column 17: | PROFDAY |
numeric; day of profile |
| column 18: | PROFTIME |
numeric; time within PROFDAY |
| column 19: | CYCLE |
numeric; count of drug administrations received |
Model generated PK and PD data to mimic an orally administered small molecule with various endpoints from continuous to ordinal response and count data. Simulated multiple dose administration ranging from 100 mg to 1600 mg, once per day.
mad_missing_duplicatesmad_missing_duplicates
A data frame with the following 19 columns:
| column 1: | ID |
numeric; unique subject ID |
| column 2: | TIME |
numeric; time relative to first drug administration |
| column 3: | NOMTIME |
numeric; nominal time |
| column 4: | TIMEUNIT |
character; unit of TIME |
| column 5: | AMT |
numeric; dosing amount (for dosing events) in mg |
| column 6: | LIDV |
numeric; observation on a linear scale (observation type determined by CMT), units determined by EVENTU column |
| column 7: | MDV |
numeric; missing dependent variable |
| column 8: | CMT |
integer; compartment number (determines observation type): |
| CMT 1 = Dosing event | ||
| CMT 2 = PK concentration | ||
| CMT 3 = Continuous response data | ||
| CMT 4 = Count response data | ||
| CMT 5 = Ordinal response data | ||
| CMT 6 = Binary response data | ||
| column 9: | NAME |
character; description of event |
| column 10: | EVENTU |
character; unit for observation |
| column 11: | CENS |
integer; censored values (0 = not censored, 1 = censored) |
| column 12: | EVID |
integer; event ID (0 = observation, 1 = dosing event) |
| column 13: | WEIGHTB |
numeric; baseline body weight (kg) |
| column 14: | SEX |
character; sex |
| column 15: | TRTACT |
factor; treatment group label |
| column 16: | DOSE |
numeric; randomized dose in mg |
| column 17: | PROFDAY |
numeric; day of profile |
| column 18: | PROFTIME |
numeric; time within PROFDAY |
| column 19: | CYCLE |
numeric; count of drug administrations received |
Multiple Ascending Dose Noncompartmental Analysis (NCA) dataset
mad_ncamad_nca
A data frame with the following 7 columns:
| column 1: | ID |
numeric; unique subject ID |
| column 2: | PARAM |
character; NCA parameter |
| column 3: | VALUE |
numeric; Value of the NCA parameter |
| column 4: | DOSE |
numeric; randomized dose in mg |
| column 15: | TRTACT |
factor; treatment group label |
| column 14: | SEX |
character; sex |
| column 13: | WEIGHTB |
numeric; baseline body weight (kg) |
Theophylline dataset, from the nlmixr R package
nlmixr_theo_sdnlmixr_theo_sd
A data frame with the following 7 columns:
| column 1: | ID |
integer; unique patient identifier |
| column 2: | TIME |
numeric; time relative to first drug administration |
| column 3: | DV |
numeric; dependent variable (drug concentration) |
| column 4: | AMT |
numeric; dose of drug |
| column 5: | EVID |
integer; event ID, 1 if dose, 0 otherwise |
| column 6: | CMT |
integer; compartment number |
| column 7: | WT |
numeric; weight |
predict.nls
## S3 method for class 'nls' predict( object, newdata = NULL, se.fit = FALSE, interval = "none", level = 0.95, ... )## S3 method for class 'nls' predict( object, newdata = NULL, se.fit = FALSE, interval = "none", level = 0.95, ... )
object |
Object of class inheriting from "nls" |
newdata |
An optional data frame in which to look for variables with which to predict. If omitted, the fitted values are used. |
se.fit |
A switch indicating if standard errors are required. |
interval |
Type of interval calculation, "none" or "confidence" |
level |
Level of confidence interval to use |
... |
additional arguments affecting the predictions produced. |
predict.nls produces a vector of predictions or a matrix of predictions and
bounds with column names fit, lwr, and upr if interval is set.
If se.fit is TRUE, a list with the following components is returned:
fit |
vector or matrix as above |
se.fit |
standard error of predicted means |
residual.scale |
residual standard deviations |
df |
degrees of freedom for residual |
set.seed(12345) data_to_plot <- data.frame(x1 = rep(c(0, 25, 50, 100, 200, 400, 600), 10)) %>% dplyr::mutate(AUC = x1*rlnorm(length(x1), 0, 0.3), x2 = x1*stats::rlnorm(length(x1), 0, 0.3), Response = (15 + 50*x2/(20+x2))*stats::rlnorm(length(x2), 0, 0.3)) gg <- ggplot2::ggplot(data = data_to_plot, ggplot2::aes(x = AUC, y = Response)) + ggplot2::geom_point() + xgx_geom_smooth(method = "nls", method.args = list(formula = y ~ E0 + Emax* x / (EC50 + x), start = list(E0 = 15, Emax = 50, EC50 = 20) ), color = "black", size = 0.5, alpha = 0.25) gg mod <- stats::nls(formula = Response ~ E0 + Emax * AUC / (EC50 + AUC), data = data_to_plot, start = list(E0 = 15, Emax = 50, EC50 = 20)) predict(mod) predict(mod, se.fit = TRUE) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE, interval = "confidence", level = 0.95) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE, interval = "confidence", level = 0.95)set.seed(12345) data_to_plot <- data.frame(x1 = rep(c(0, 25, 50, 100, 200, 400, 600), 10)) %>% dplyr::mutate(AUC = x1*rlnorm(length(x1), 0, 0.3), x2 = x1*stats::rlnorm(length(x1), 0, 0.3), Response = (15 + 50*x2/(20+x2))*stats::rlnorm(length(x2), 0, 0.3)) gg <- ggplot2::ggplot(data = data_to_plot, ggplot2::aes(x = AUC, y = Response)) + ggplot2::geom_point() + xgx_geom_smooth(method = "nls", method.args = list(formula = y ~ E0 + Emax* x / (EC50 + x), start = list(E0 = 15, Emax = 50, EC50 = 20) ), color = "black", size = 0.5, alpha = 0.25) gg mod <- stats::nls(formula = Response ~ E0 + Emax * AUC / (EC50 + AUC), data = data_to_plot, start = list(E0 = 15, Emax = 50, EC50 = 20)) predict(mod) predict(mod, se.fit = TRUE) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE, interval = "confidence", level = 0.95) predict(mod, newdata = data.frame(AUC = c(0, 25, 50, 100, 200, 400, 600)), se.fit = TRUE, interval = "confidence", level = 0.95)
Prediction data frame from ggplot2 Get predictions with standard errors into data frame
predictdf(model, xseq, se, level)predictdf(model, xseq, se, level)
model |
model object |
xseq |
newdata |
se |
Display confidence interval around smooth? |
level |
Level of confidence interval to use |
Get predictions with standard errors into data frame for use with geom_smooth
## S3 method for class 'nls' predictdf(model, xseq, se, level)## S3 method for class 'nls' predictdf(model, xseq, se, level)
model |
nls object |
xseq |
newdata |
se |
Display confidence interval around smooth? |
level |
Level of confidence interval to use |
ggplot2::geom_smooth produces confidence intervals by silently calling functions
of the form predictdf.method, where method is "loess", "lm", "glm" etc.
depending on what method is specified in the call to geom_smooth.
Currently ggplot2 does not define a predictdf.nls function for method of type "nls",
and thus confidence intervals cannot be automatically generated by geom_smooth
for method = "nls". Here we define predictdf.nls for calculating the confidence
intervals of an object of type nls. geom_smooth will silently call this function
whenever method = "nls", and produce the appropriate confidence intervals.
predictdf.nls calculates CI for a model fit of class nls based on the "delta-method"
http://sia.webpopix.org/nonlinearRegression.html#confidence-intervals-and-prediction-intervals)
CI = [ f(x0, beta) + qt_(alpha/2, n - d) * se(f(x0, beta)), f(x0, beta) + qt_(1 - alpha/2, n - d) * se(f(x0, beta))]
where: beta = vector of parameter estimates x = independent variable se(f(x0, beta)) = sqrt( delta(f)(x0, beta) * Var(beta) * (delta(f)(x0, beta))' ) delta(f) is the gradient of f
dataframe with x and y values, if se is TRUE dataframe also includes ymin and ymax
Get predictions with standard errors into data frame for use with geom_smooth
## S3 method for class 'polr' predictdf(model, xseq, se, level)## S3 method for class 'polr' predictdf(model, xseq, se, level)
model |
object returned from polr |
xseq |
sequence of x values for which to compute the smooth |
se |
if TRUE then confidence intervals are returned |
level |
confidence level for confidence intervals |
predictdf.polr is used by xgx_geom_smooth when method = "polr"
to calculate confidence intervals via bootstraps.
Model generated PK data to mimic an orally administered small molecule. Simulated single dose administration ranging from 100 mg to 1600 mg.
sadsad
A data frame with the following 16 columns:
| column 1: | ID |
numeric; unique subject ID |
| column 2: | TIME |
numeric; time relative to first drug administration |
| column 3: | NOMTIME |
numeric; nominal time |
| column 4: | TIMEUNIT |
character; unit of TIME |
| column 5: | AMT |
numeric; dosing amount (for dosing events) in mg |
| column 6: | LIDV |
numeric; observation on a linear scale (observation type determined by CMT), units determined by EVENTU column |
| column 7: | MDV |
numeric; missing dependent variable |
| (1 if missing, 0 otherwise) | ||
| column 8: | CMT |
integer; compartment number (determines observation type): |
| CMT 1 = Dosing event | ||
| CMT 2 = PK concentration | ||
| column 9: | NAME |
character; description of event |
| column 10: | EVENTU |
character; unit for observation |
| column 11: | CENS |
integer; censored values (0 = not censored, 1 = censored) |
| column 12: | EVID |
integer; event ID (0 = observation, 1 = dosing event) |
| column 13: | WEIGHTB |
numeric; baseline body weight (kg) |
| column 14: | SEX |
character; sex |
| column 15: | TRTACT |
factor; treatment group label |
| column 16: | DOSE |
numeric; randomized dose in mg received |
Stat object for producing smooths through ordinal data
StatSmoothOrdinalStatSmoothOrdinal
An object of class StatSmoothOrdinal (inherits from Stat, ggproto, gg) of length 8.
Source: https://github.com/tidyverse/ggplot2/blob/351eb41623397dea20ed0059df62a4a5974d88cb/R/stat-summary-bin.R
StatSummaryBinQuantStatSummaryBinQuant
An object of class StatSummaryBinQuant (inherits from Stat, ggproto, gg) of length 5.
StatSummaryBinQuant returns a ggproto object for plotting mean +/- confidence bins
ggplot2 ggproto object
StatSummaryOrdinal returns a ggproto object for plotting mean +/- confidence intervals
for ordinal data. It also allows for binning values on the independent axis.
StatSummaryOrdinalStatSummaryOrdinal
An object of class StatSummaryOrdinal (inherits from Stat, ggproto, gg) of length 8.
ggplot2 ggproto object
Calls the standard theme for xGx graphics
theme_xgx()theme_xgx()
xgx ggplot2 compatible theme
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10() + theme_xgx()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10() + theme_xgx()
xgx_annotate_filenames appends file details to the bottom of a plot
using the plot caption option.
File details to append include the parent directory, the path of the R
script which generated the plot,
and the path of the plot.
xgx_annotate_filenames(dirs, hjust = 0.5, color = "black", size = 11)xgx_annotate_filenames(dirs, hjust = 0.5, color = "black", size = 11)
dirs |
list containing directories and filenames. It must contain five fields
|
hjust |
horizontal justification of the caption |
color |
font color for caption, default black |
size |
font size for caption, default 11 |
None
dirs <- list(parent_dir = "/your/parent/path/", rscript_dir = "./Rscripts/", rscript_name = "Example.R", results_dir = "./Results/", filename = "your_file_name.png") data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_filenames(dirs)dirs <- list(parent_dir = "/your/parent/path/", rscript_dir = "./Rscripts/", rscript_name = "Example.R", results_dir = "./Results/", filename = "your_file_name.png") data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_filenames(dirs)
xgx_annotate_status adds a status (e.g. DRAFT) annotation layer
to a plot.
The text of the annotation can be customized, the default is "DRAFT".
The color, location, size, fontface, transparency of the annotation can
also be customized.
xgx_annotate_status( status = "DRAFT", x = Inf, y = Inf, color = "grey", hjust = 1.2, vjust = 1.2, fontsize = 7, fontface = "bold", alpha = 0.5, ... )xgx_annotate_status( status = "DRAFT", x = Inf, y = Inf, color = "grey", hjust = 1.2, vjust = 1.2, fontsize = 7, fontface = "bold", alpha = 0.5, ... )
status |
the text to |
x |
x location, default Inf (right most point) |
y |
y location, default Inf (up most point) |
color |
font color, default "grey" |
hjust |
horizontal justification, default 1.2 |
vjust |
vertical justification, default 1.2 |
fontsize |
font size to use, default 7 |
fontface |
font style to use, default "bold" |
alpha |
transparency, default is 0.5 |
... |
other arguments passed on to |
The default value of x = Inf works in many instances, but not some special data types like Dates. In that case, you will need to specify x
ggplot layer
data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_status("DRAFT") # for dates in x-axis data <- data.frame(x = as.Date(c("2015-01-01", "2016-01-01")), y = c(1,2)) xmax <- max(na.omit(data$x)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_status("DRAFT", x = xmax)data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_status("DRAFT") # for dates in x-axis data <- data.frame(x = as.Date(c("2015-01-01", "2016-01-01")), y = c(1,2)) xmax <- max(na.omit(data$x)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_annotate_status("DRAFT", x = xmax)
These function annotates a single png file or all files within a directory.
xgx_annotate_status_png( file_or_dir, script = "", status = "DRAFT", date_format = "%a %b %d %X %Y", col = grDevices::grey(0.8, alpha = 0.7), font = 2, cex_status_mult = 7, cex_footnote_mult = 0.8, status_angle = 45, x11 = FALSE )xgx_annotate_status_png( file_or_dir, script = "", status = "DRAFT", date_format = "%a %b %d %X %Y", col = grDevices::grey(0.8, alpha = 0.7), font = 2, cex_status_mult = 7, cex_footnote_mult = 0.8, status_angle = 45, x11 = FALSE )
file_or_dir |
The png file to annotate or directory location for annotating png files. Note this will annotate just once, so if you generate multiple png files and then annotate at the end of your script it will have the correct script name on it. Then if you create new images in a different script in the same directory and then annotate with the script name the second script, the PNG files will show the correct script location for each file. |
script |
Script name to add as a footnote; By default this is empty, though it could name the script that |
status |
Draft or other status; If |
date_format |
Date format for adding the time the png was annotated. |
col |
Color for annotating the draft status |
font |
Font to use for the annotation function |
cex_status_mult |
Multiplication factor for the status annotation. By default 7 |
cex_footnote_mult |
Multiplication factor for the footnote annotation. By default 0.8 |
status_angle |
Angle to rotate status |
x11 |
Display on the X11/Windows device |
If a png file has been annotated once, this function will not annotate it again. Therefore, you can run this function on directories with different input script names and it will label each file based on when each file was run.
Based on code from MrFlick on Stack Overflow.
nothing
Matthew Fidler, Alison M, ....
# using the examples from plot() file.name <- tempfile() grDevices::png(file.name) graphics::plot(cars) graphics::lines(stats::lowess(cars)) grDevices::dev.off() # annotate one file xgx_annotate_status_png(file.name, "/tmp/script1.R")# using the examples from plot() file.name <- tempfile() grDevices::png(file.name) graphics::plot(cars) graphics::lines(stats::lowess(cars)) grDevices::dev.off() # annotate one file xgx_annotate_status_png(file.name, "/tmp/script1.R")
xgx_auto_explore returns an HTML and PDF document with plots
describing the provided dataset
xgx_auto_explore( data_path = NULL, mapping = list(), author_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL, dose_cmt = NULL, steady_state_day = NULL, time_between_doses = NULL, rmd_template_name = NULL, rmd_template_path = NULL, rmd_output_path = NULL, pdf_output_path = NULL, html_output_path = NULL, add_datetime = TRUE, show_explanation = TRUE )xgx_auto_explore( data_path = NULL, mapping = list(), author_name = NULL, multiple_dosing = FALSE, pk_cmt = NULL, pd_cmt = NULL, pd_data_type = NULL, dose_cmt = NULL, steady_state_day = NULL, time_between_doses = NULL, rmd_template_name = NULL, rmd_template_path = NULL, rmd_output_path = NULL, pdf_output_path = NULL, html_output_path = NULL, add_datetime = TRUE, show_explanation = TRUE )
data_path |
Path (as a string) to the dataset that is to be analyzed |
mapping |
A list of column name mappings from the original (template) dataset column names to the corresponding columns in the new dataset. |
author_name |
The name of the author to be displayed on the template |
multiple_dosing |
Whether or not to use a "Multiple" or "Single" Ascending dose template |
pk_cmt |
An integer denoting the "compartment" containing the PK data. The "CMT" column will typically have these integers, where each row may contain PK, PD, dosing or other events/observations data |
pd_cmt |
An integer denoting the "compartment" containing the PD data, of the desired type (continuous, ordinal, etc.). The "CMT" column will typically have these integers, where each row may contain PK, PD, dosing or other events/observations data |
pd_data_type |
The type of PD data - acceptable values exist in the following list: ["binary","continuous","count","ordinal","real_example","receptor_occupancy","time_to_event"] |
dose_cmt |
Integer denoting the compartment for dosing records |
steady_state_day |
used to denote the day of rich sampling of PK at steady state |
time_between_doses |
dosing interval, has units to match the time variable of the dataset |
rmd_template_name |
A custom output name for the generated Rmd file |
rmd_template_path |
A user provided custom template (as a string) |
rmd_output_path |
A custom output path for the generated Rmd file (This is typically left as 'NULL' in order to maintain the hierarchical directory structure of 'xgx_autoexplore_output')) |
pdf_output_path |
A custom output path for the generated PDF file (This is typically left as 'NULL' in order to maintain the hierarchical directory structure of 'xgx_autoexplore_output')) |
html_output_path |
A custom output path for the generated HTML file (This is typically left as 'NULL' in order to maintain the hierarchical directory structure of 'xgx_autoexplore_output')) |
add_datetime |
Boolean indicating additon of a date stamp to the beginnning of the Rmd file |
show_explanation |
Boolean indicating if the additional explanations (text in between figures) are needed for the user. |
This function can be used quickly to explore your data by generating overview plots before constructing non-linear mixed effects models.
author_name = "Your Name Here" show_explanation = FALSE ## Not run: # Try out the nonlinear_pkpd dataset with the # Multiple Ascending Dose PK Rmd template data_path <- "~/nonlinear_pkpd.csv" # Specify the mapping of column names mapping <- list( "TIME" = "TIM2", "NOMTIME" = "NT", "EVID" = 0, "CENS" = 0, "DOSE" = "MGKG", "TRTACT" = "TRT", "LIDV_NORM" = "LIDV/MGKG", "LIDV_UNIT" = "UNIT", "PROFDAY" = 1, "SEX" = 0, "WEIGHTB" = 0) # 5 contains the PK Concentration in this dataset pk_cmt = 5 # We don't need PD right now pd_cmt = NULL pd_data_type = NULL dose_cmt = 1 steady_state_day = c(0, 6) time_between_doses = 24 multiple_dosing = TRUE output_directory = tempdir() xgx_auto_explore(data_path = data_path, mapping = mapping, author_name = author_name, pk_cmt = pk_cmt, pd_cmt = pd_cmt, dose_cmt = dose_cmt, steady_state_day = steady_state_day, time_between_doses = time_between_doses, multiple_dosing = multiple_dosing, pd_data_type = pd_data_type, rmd_output_path = output_directory, show_explanation = show_explanation) ## End(Not run)author_name = "Your Name Here" show_explanation = FALSE ## Not run: # Try out the nonlinear_pkpd dataset with the # Multiple Ascending Dose PK Rmd template data_path <- "~/nonlinear_pkpd.csv" # Specify the mapping of column names mapping <- list( "TIME" = "TIM2", "NOMTIME" = "NT", "EVID" = 0, "CENS" = 0, "DOSE" = "MGKG", "TRTACT" = "TRT", "LIDV_NORM" = "LIDV/MGKG", "LIDV_UNIT" = "UNIT", "PROFDAY" = 1, "SEX" = 0, "WEIGHTB" = 0) # 5 contains the PK Concentration in this dataset pk_cmt = 5 # We don't need PD right now pd_cmt = NULL pd_data_type = NULL dose_cmt = 1 steady_state_day = c(0, 6) time_between_doses = 24 multiple_dosing = TRUE output_directory = tempdir() xgx_auto_explore(data_path = data_path, mapping = mapping, author_name = author_name, pk_cmt = pk_cmt, pd_cmt = pd_cmt, dose_cmt = dose_cmt, steady_state_day = steady_state_day, time_between_doses = time_between_doses, multiple_dosing = multiple_dosing, pd_data_type = pd_data_type, rmd_output_path = output_directory, show_explanation = show_explanation) ## End(Not run)
xgx_breaks_log10 sets nice breaks for log10 scale.
it's better than the default function because it ensures there is at least
2 breaks
and also, it will try to go by 3s (i.e. 1,3,10,30,100) if it makes sense
xgx_breaks_log10(data_range)xgx_breaks_log10(data_range)
data_range |
range of the data |
for the extended breaks function, weights is a set of 4 weights for
simplicity - how early in the Q order are you
coverage - labelings that don't extend outside the data: range(data) / range(labels)
density (previously granularity) - how close to the number of ticks do you get (default is 5)
legibility - has to do with fontsize and formatting to prevent label overlap
numeric vector of breaks
Talbot, Justin, Sharon Lin, and Pat Hanrahan. "An extension of Wilkinson’s algorithm for positioning tick labels on axes." IEEE Transactions on visualization and computer graphics 16.6 (2010): 1036-1043.
xgx_breaks_log10(c(1, 1000)) xgx_breaks_log10(c(0.001, 100)) xgx_breaks_log10(c(1e-4, 1e4)) xgx_breaks_log10(c(1e-9, 1e9)) xgx_breaks_log10(c(1, 2)) xgx_breaks_log10(c(1, 5)) xgx_breaks_log10(c(1, 10)) xgx_breaks_log10(c(1, 100)) xgx_breaks_log10(c(1, 1.01)) xgx_breaks_log10(c(1, 1.0001)) print(xgx_breaks_log10(c(1, 1.000001)), digits = 10)xgx_breaks_log10(c(1, 1000)) xgx_breaks_log10(c(0.001, 100)) xgx_breaks_log10(c(1e-4, 1e4)) xgx_breaks_log10(c(1e-9, 1e9)) xgx_breaks_log10(c(1, 2)) xgx_breaks_log10(c(1, 5)) xgx_breaks_log10(c(1, 10)) xgx_breaks_log10(c(1, 100)) xgx_breaks_log10(c(1, 1.01)) xgx_breaks_log10(c(1, 1.0001)) print(xgx_breaks_log10(c(1, 1.000001)), digits = 10)
xgx_breaks_time sets the default breaks for a time axis,
given the units of the data and the units of the plot.
It is inspired by scales::extended_breaks
xgx_breaks_time(data_range, units_plot, number_breaks = 5)xgx_breaks_time(data_range, units_plot, number_breaks = 5)
data_range |
range of the data |
units_plot |
units to use in the plot |
number_breaks |
number of breaks to aim for (default is 5) |
for the extended breaks function, weights is a set of 4 weights for
simplicity - how early in the Q order are you
coverage - labelings that don't extend outside the data: range(data) / range(labels)
density (previously granularity) - how close to the number of ticks do you get (default is 5)
legibility - has to do with fontsize and formatting to prevent label overlap
numeric vector of breaks
Talbot, Justin, Sharon Lin, and Pat Hanrahan. "An extension of Wilkinson’s algorithm for positioning tick labels on axes." IEEE Transactions on visualization and computer graphics 16.6 (2010): 1036-1043.
xgx_breaks_time(c(0, 5), "hour") xgx_breaks_time(c(0, 6), "hour") xgx_breaks_time(c(-3, 5), "hour") xgx_breaks_time(c(0, 24), "hour") xgx_breaks_time(c(0, 12), "hour") xgx_breaks_time(c(1, 4), "day") xgx_breaks_time(c(1, 12), "day") xgx_breaks_time(c(1, 14), "day") xgx_breaks_time(c(1, 50), "day") xgx_breaks_time(c(1000, 3000), "day") xgx_breaks_time(c(-21, 100), "day") xgx_breaks_time(c(-1, 10), "week")xgx_breaks_time(c(0, 5), "hour") xgx_breaks_time(c(0, 6), "hour") xgx_breaks_time(c(-3, 5), "hour") xgx_breaks_time(c(0, 24), "hour") xgx_breaks_time(c(0, 12), "hour") xgx_breaks_time(c(1, 4), "day") xgx_breaks_time(c(1, 12), "day") xgx_breaks_time(c(1, 14), "day") xgx_breaks_time(c(1, 50), "day") xgx_breaks_time(c(1000, 3000), "day") xgx_breaks_time(c(-21, 100), "day") xgx_breaks_time(c(-1, 10), "week")
xgx_check_data performs a series of checks on a PK or PKPD dataset
It was inspired by the dataset preparation table from
IntiQuan.
xgx_check_data(data, covariates = NULL)xgx_check_data(data, covariates = NULL)
data |
the dataset to check. Must contain the above columns |
covariates |
the column names of covariates, to explore |
The dataset must have the following columns
ID = unique subject identifier. USUBJID is another option if ID is not there
EVID = event ID: 1 for dose, 0 otherwise
AMT = value of the dose
TIME = time of the measurement
DV = dependent value (linear scale). will check if LIDV or LNDV are also there if DV is not
YTYPE = data measurement for LIDV. will check if CMT is there, if YTYPE is not
The dataset may also have additional columns
CENS = flag for censoring of the data because it's below the limit of quantification (BLOQ)
MDV = missing dependent variable - will be counted and then filtered out from the data check
data.frame
covariates <- c("WEIGHTB", "SEX") check <- xgx_check_data(mad_missing_duplicates, covariates)covariates <- c("WEIGHTB", "SEX") check <- xgx_check_data(mad_missing_duplicates, covariates)
xgx_conf_int returns a dataframe with mean +/- confidence intervalsxgx_conf_int returns a dataframe with mean +/- confidence intervals
xgx_conf_int(y, conf_level = 0.95, distribution = "normal", ci_method = NULL)xgx_conf_int(y, conf_level = 0.95, distribution = "normal", ci_method = NULL)
y |
data to compute confidence interval of |
conf_level |
The percentile for the confidence interval (should fall between 0 and 1). The default is 0.95, which corresponds to a 95 percent confidence interval. |
distribution |
The distribution which the data follow, used for
calculating confidence intervals. The options are "normal", "lognormal",
and "binomial". The "normal" option will use the Student t Distribution
to calculate confidence intervals, the "lognormal" option will transform
data to the log space first. The "binomial" option will use the
|
ci_method |
Method to pass to binom.confint or MultinomCI. Defaults are "exact" and "goodman", respectively. |
data.frame
# default settings for normally distributed data, 95% confidence interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_conf_int(data$y)# default settings for normally distributed data, 95% confidence interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_conf_int(data$y)
xgx_dirs2char returns a character variable based on the dirs list.
The caption gives the filename
xgx_dirs2char(dirs, include_time = TRUE)xgx_dirs2char(dirs, include_time = TRUE)
dirs |
list containing directories and filenames. It must contain five fields
|
include_time |
is logical with default TRUE. If TRUE, it includes date / time in the output character |
character
dirs <- list(parent_dir = "/your/parent/path/", rscript_dir = "./Rscripts/", rscript_name = "Example.R", results_dir = "./Results/", filename = "your_file_name.png") caption <- xgx_dirs2char(dirs)dirs <- list(parent_dir = "/your/parent/path/", rscript_dir = "./Rscripts/", rscript_name = "Example.R", results_dir = "./Results/", filename = "your_file_name.png") caption <- xgx_dirs2char(dirs)
Plot data with mean and confidence intervals
xgx_geom_ci( mapping = NULL, data = NULL, conf_level = 0.95, distribution = "normal", bins = NULL, breaks = NULL, geom = list("point", "line", "errorbar"), position = "identity", fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )xgx_geom_ci( mapping = NULL, data = NULL, conf_level = 0.95, distribution = "normal", bins = NULL, breaks = NULL, geom = list("point", "line", "errorbar"), position = "identity", fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )
mapping |
Set of aesthetic mappings created by 'aes' or 'aes_'. If specified and 'inherit.aes = TRUE' (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping. |
data |
The data to be displayed in this layer. There are three options: If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot. A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify for which variables will be created. A function will be called with a single argument, the plot data. The return value must be a data.frame., and will be used as the layer data. |
conf_level |
The percentile for the confidence interval (should fall between 0 and 1). The default is 0.95, which corresponds to a 95 percent confidence interval. |
distribution |
The distribution which the data follow, used for
calculating confidence intervals. The options are "normal", "lognormal",
and "binomial". The "normal" option will use the Student t Distribution
to calculate confidence intervals, the "lognormal" option will transform
data to the log space first. The "binomial" option will use the
|
bins |
number of bins to cut up the x data, cuts data into quantiles. |
breaks |
breaks to cut up the x data, if this option is used, bins is ignored |
geom |
Use to override the default geom. Can be a list of multiple geoms, e.g. list("point","line","errorbar"), which is the default. |
position |
Position adjustment, either as a string, or the result of a call to a position adjustment function. |
fun.args |
Optional additional arguments passed on to the functions. |
na.rm |
If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed. |
show.legend |
logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. |
inherit.aes |
If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders. |
... |
other arguments passed on to layer. These are often aesthetics, used to set an aesthetic to a fixed value, like color = "red" or size = 3. They may also be parameters to the paired geom/stat. |
ggplot2 plot layer
data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + xgx_geom_ci(conf_level = 0.95)data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + xgx_geom_ci(conf_level = 0.95)
Plot data with median and percent intervals
xgx_geom_pi( mapping = NULL, data = NULL, percent_level = 0.95, geom = list("line", "ribbon"), position = "identity", fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )xgx_geom_pi( mapping = NULL, data = NULL, percent_level = 0.95, geom = list("line", "ribbon"), position = "identity", fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )
mapping |
Set of aesthetic mappings created by 'aes' or 'aes_'. If specified and 'inherit.aes = TRUE' (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping. |
data |
The data to be displayed in this layer. There are three options: If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot. A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify for which variables will be created. A function will be called with a single argument, the plot data. The return value must be a data.frame., and will be used as the layer data. |
percent_level |
The upper or lower percentile for the percent interval (should fall between 0 and 1). The default is 0.95, which corresponds to (0.05, 0.95) interval. Supplying 0.05 would give the same result |
geom |
Use to override the default geom. Can be a list of multiple geoms, e.g. list("line","ribbon"), which is the default. |
position |
Position adjustment, either as a string, or the result of a call to a position adjustment function. |
fun.args |
Optional additional arguments passed on to the functions. |
na.rm |
If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed. |
show.legend |
logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. |
inherit.aes |
If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders. |
... |
other arguments passed on to layer. These are often aesthetics, used to set an aesthetic to a fixed value, like color = "red" or size = 3. They may also be parameters to the paired geom/stat. |
ggplot2 plot layer
data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + xgx_geom_pi(percent_level = 0.95)data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60)) ggplot2::ggplot(data, ggplot2::aes(x = x, y = y)) + xgx_geom_pi(percent_level = 0.95)
Returns a set of labels for ggplot
xgx_labels_log10(breaks)xgx_labels_log10(breaks)
breaks |
breaks for the function |
either character or expression
print(xgx_labels_log10(c(1e-5, 1, 1e5)))print(xgx_labels_log10(c(1e-5, 1, 1e5)))
xgx_minor_breaks_log10 sets nice minor_breaks for log10 scale.
xgx_minor_breaks_log10(data_range)xgx_minor_breaks_log10(data_range)
data_range |
range of the data |
numeric vector of breaks
xgx_minor_breaks_log10(c(1, 1000)) xgx_minor_breaks_log10(c(0.001, 100)) xgx_minor_breaks_log10(c(1e-4, 1e4)) xgx_minor_breaks_log10(c(1e-9, 1e9)) xgx_minor_breaks_log10(c(1, 2)) xgx_minor_breaks_log10(c(1, 5)) xgx_minor_breaks_log10(c(1, 10)) xgx_minor_breaks_log10(c(1, 100)) xgx_minor_breaks_log10(c(1, 1.01)) xgx_minor_breaks_log10(c(1, 1.0001)) print(xgx_minor_breaks_log10(c(1, 1.000001)), digits = 10)xgx_minor_breaks_log10(c(1, 1000)) xgx_minor_breaks_log10(c(0.001, 100)) xgx_minor_breaks_log10(c(1e-4, 1e4)) xgx_minor_breaks_log10(c(1e-9, 1e9)) xgx_minor_breaks_log10(c(1, 2)) xgx_minor_breaks_log10(c(1, 5)) xgx_minor_breaks_log10(c(1, 10)) xgx_minor_breaks_log10(c(1, 100)) xgx_minor_breaks_log10(c(1, 1.01)) xgx_minor_breaks_log10(c(1, 1.0001)) print(xgx_minor_breaks_log10(c(1, 1.000001)), digits = 10)
Create a new xgx plot
xgx_plot( data = NULL, mapping = ggplot2::aes(), ..., environment = parent.frame() )xgx_plot( data = NULL, mapping = ggplot2::aes(), ..., environment = parent.frame() )
data |
Default dataset to use for plot. If not already a data.frame, will be converted to one by fortify. |
mapping |
As in ggplot2; Default list of aesthetic mappings to use for plot. Must define x, y, and group for xgx_spaghetti. |
... |
Other arguments passed on to methods. Not currently used. |
environment |
If an variable defined in the aesthetic mapping is not
found in the data, ggplot will look for it in this environment. It defaults
to using the environment in which |
ggplot2 object
time <- rep(seq(1, 10), 5) id <- sort(rep(seq(1, 5), 10)) conc <- exp(-time) * sort(rep(stats::rlnorm(5), 10)) data <- data.frame(time = time, concentration = conc, id = id) xgx_plot(data = data, mapping = ggplot2::aes(x = time, y = concentration, group = id)) + ggplot2::geom_line() + ggplot2::geom_point()time <- rep(seq(1, 10), 5) id <- sort(rep(seq(1, 5), 10)) conc <- exp(-time) * sort(rep(stats::rlnorm(5), 10)) data <- data.frame(time = time, concentration = conc, id = id) xgx_plot(data = data, mapping = ggplot2::aes(x = time, y = concentration, group = id)) + ggplot2::geom_line() + ggplot2::geom_point()
Saving plot, automatically annotating the status and denoting the filenames
xgx_save( width, height, dirs = NULL, filename_main = NULL, status = "DRAFT", g = ggplot2::last_plot(), filetype = "png", status_x = Inf, status_y = Inf, status_fontsize = 7, status_fontcolor = "grey", filenames_fontsize = 11, filenames_fontcolor = "black" )xgx_save( width, height, dirs = NULL, filename_main = NULL, status = "DRAFT", g = ggplot2::last_plot(), filetype = "png", status_x = Inf, status_y = Inf, status_fontsize = 7, status_fontcolor = "grey", filenames_fontsize = 11, filenames_fontcolor = "black" )
width |
width of plot |
height |
height of plot |
dirs |
list of directories. If NULL or if directories missing, there is default behavior below
|
filename_main |
main part of the filename, excluding prefix and suffix. no default |
status |
status to be annotated |
g |
ggplot plot object, default is ggplot::last_plot() |
filetype |
file extension (e.g. "pdf","csv" etc.) |
status_x |
x location of the status in plot |
status_y |
y location of the status in plot |
status_fontsize |
font size for status in plot |
status_fontcolor |
font color for status in plot |
filenames_fontsize |
font size for filenames info in plot |
filenames_fontcolor |
font color for filenames info in plot |
ggplot2 plot object
directory = tempdir() dirs <- list(parent_dir = directory, rscript_dir = directory, rscript_name = "example.R", results_dir = directory, filename_prefix = "example_") data <- data.frame(x = 1:1000, y = stats::rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() xgx_save(4, 4, dirs, "Example", "DRAFT")directory = tempdir() dirs <- list(parent_dir = directory, rscript_dir = directory, rscript_name = "example.R", results_dir = directory, filename_prefix = "example_") data <- data.frame(x = 1:1000, y = stats::rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() xgx_save(4, 4, dirs, "Example", "DRAFT")
Saving table as an image, also labeling the program that created the table and where the table is stored
xgx_save_table(data, dirs = NULL, filename_main = NULL)xgx_save_table(data, dirs = NULL, filename_main = NULL)
data |
data.frame or table of results |
dirs |
list of directories. If NULL or if directories missing, there is default behavior below
|
filename_main |
main part of the filename, excluding prefix and extension. no default |
ggplot2 plot object
directory = tempdir() dirs <- list(parent_dir = directory, rscript_dir = directory, rscript_name = "example.R", results_dir = directory, filename_prefix = "example_") data <- data.frame(x = c(1, 2), y = c(1, 2)) xgx_save_table(data, dirs = dirs, filename_main = "test")directory = tempdir() dirs <- list(parent_dir = directory, rscript_dir = directory, rscript_name = "example.R", results_dir = directory, filename_prefix = "example_") data <- data.frame(x = c(1, 2), y = c(1, 2)) xgx_save_table(data, dirs = dirs, filename_main = "test")
xgx_scale_x_log10 is similar to
scale_x_log10.
But it uses what we believe to be a nicer spacing and set of tick marks
it can be used the same as
scale_x_log10
xgx_scale_x_log10( breaks = xgx_breaks_log10, minor_breaks = NULL, labels = xgx_labels_log10, ... )xgx_scale_x_log10( breaks = xgx_breaks_log10, minor_breaks = NULL, labels = xgx_labels_log10, ... )
breaks |
major breaks, default is a function defined here |
minor_breaks |
minor breaks, default is a function defined here |
labels |
function for setting the labels, defined here |
... |
other arguments passed to
|
ggplot2 compatible scale object
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(x = concentration, y = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_x_log10() + xgx_scale_y_reverselog10()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(x = concentration, y = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_x_log10() + xgx_scale_y_reverselog10()
xgx_scale_x_reverselog10 is designed to be used with data that
approaches 100
A common example is receptor occupancy in drug development.
It is used when you want even spacing between 90, 99, 99.9, etc.
xgx_scale_x_reverselog10(labels = NULL, accuracy = NULL, ...)xgx_scale_x_reverselog10(labels = NULL, accuracy = NULL, ...)
labels |
if NULL, then the default is to use scales::percent() |
accuracy |
if NULL, then use the the default as specified by scales::percent() to round to the hundredths place, set accuracy 0.01 |
... |
other parameters passed to
|
ggplot2 compatible scale object
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()
xgx_scale_x_time_units converts x axis scale from one time unit
to another.
Supported units include hours, days, weeks, months, and years, which
can also be called using just the first letter (h, d, w, m, y).
xgx_scale_x_time_units( units_dataset, units_plot = NULL, breaks = NULL, labels = NULL, ... ) xgx_scale_y_time_units( units_dataset, units_plot = NULL, breaks = NULL, labels = NULL, ... )xgx_scale_x_time_units( units_dataset, units_plot = NULL, breaks = NULL, labels = NULL, ... ) xgx_scale_y_time_units( units_dataset, units_plot = NULL, breaks = NULL, labels = NULL, ... )
units_dataset |
units of the input dataset, must be specified by user as "h", "d", "w", "m", or "y" |
units_plot |
units of the plot, will be units of the dataset if empty |
breaks |
One of:
|
labels |
One of:
|
... |
other parameters for
|
Note: xgx_scale_x_time_units only scales the plot axis, all other
specifications must be on the original scale of the dataset (e.g. breaks,
position, width)
ggplot2 compatible scale object
data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_scale_x_time_units(units_dataset = "hours", units_plot = "weeks")data <- data.frame(x = 1:1000, y = rnorm(1000)) ggplot2::ggplot(data = data, ggplot2::aes(x = x, y = y)) + ggplot2::geom_point() + xgx_scale_x_time_units(units_dataset = "hours", units_plot = "weeks")
xgx_scale_y_log10 is similar to
scale_y_log10.
But it uses what we believe to be a nicer spacing and set of tick marks
it can be used the same as
scale_y_log10
xgx_scale_y_log10( breaks = xgx_breaks_log10, minor_breaks = NULL, labels = xgx_labels_log10, ... )xgx_scale_y_log10( breaks = xgx_breaks_log10, minor_breaks = NULL, labels = xgx_labels_log10, ... )
breaks |
major breaks, default is a function defined here |
minor_breaks |
minor breaks, default is a function defined here |
labels |
function for setting the labels, defined here |
... |
other arguments passed to
|
ggplot2 compatible scale object
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()
xgx_scale_y_percentchangelog10 and xgx_scale_x_percentchangelog10 are designed
to be used with percent change (PCHG) from baseline data (on a scale of -1 to +Inf).
Common examples include
It is used when you have a wide range of data on a percent change scale,
especially data close to -100
xgx_scale_y_percentchangelog10( breaks = NULL, minor_breaks = NULL, labels = NULL, accuracy = 1, n_breaks = 7, ... ) xgx_scale_x_percentchangelog10( breaks = NULL, minor_breaks = NULL, labels = NULL, accuracy = 1, n_breaks = 7, ... )xgx_scale_y_percentchangelog10( breaks = NULL, minor_breaks = NULL, labels = NULL, accuracy = 1, n_breaks = 7, ... ) xgx_scale_x_percentchangelog10( breaks = NULL, minor_breaks = NULL, labels = NULL, accuracy = 1, n_breaks = 7, ... )
breaks |
if NULL, then default is to use a variant of 2^(labeling::extended(log2(PCHG + 1))) - 1, where PCHG represents the range of the data |
minor_breaks |
if NULL, then default is to use nicely spaced log10(PCHG + 1) minor breaks |
labels |
if NULL, then the default is to use scales::percent_format() |
accuracy |
accuracy to use with scales::percent_format(), if NULL, then the default is set to 1 |
n_breaks |
number of desired breaks, if NULL, then the default is set to 7 |
... |
other parameters passed to
|
ggplot2 compatible scale object
dat1 <- data.frame(x = rnorm(100), PCHG = exp(rnorm(100)) - 1) ggplot2::ggplot(dat1, ggplot2::aes(x = x, y = PCHG)) + ggplot2::geom_point() + xgx_theme() + xgx_scale_y_percentchangelog10()dat1 <- data.frame(x = rnorm(100), PCHG = exp(rnorm(100)) - 1) ggplot2::ggplot(dat1, ggplot2::aes(x = x, y = PCHG)) + ggplot2::geom_point() + xgx_theme() + xgx_scale_y_percentchangelog10()
xgx_scale_y_reverselog10 is designed to be used with data
that approaches 100
A common example is receptor occupancy in drug development.
It is used when you want even spacing between 90, 99, 99.9, etc.
xgx_scale_y_reverselog10(labels = NULL, accuracy = NULL, ...)xgx_scale_y_reverselog10(labels = NULL, accuracy = NULL, ...)
labels |
if NULL, then the default is to use scales::percent() |
accuracy |
if NULL, then use the the default as specified by scales::percent() to round to the hundredths place, set accuracy 0.01 |
... |
other parameters passed to
|
ggplot2 compatible scale object
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(x = concentration, y = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_x_log10() + xgx_scale_y_reverselog10()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(x = concentration, y = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_x_log10() + xgx_scale_y_reverselog10()
xgx_stat_ci returns a ggplot layer plotting mean +/- confidence
intervals
xgx_stat_ci( mapping = NULL, data = NULL, conf_level = 0.95, distribution = "normal", bins = NULL, breaks = NULL, geom = list("point", "line", "errorbar"), position = "identity", fun.args = list(), fun.data = NULL, na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE, ... )xgx_stat_ci( mapping = NULL, data = NULL, conf_level = 0.95, distribution = "normal", bins = NULL, breaks = NULL, geom = list("point", "line", "errorbar"), position = "identity", fun.args = list(), fun.data = NULL, na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE, ... )
mapping |
Set of aesthetic mappings created by 'aes' or 'aes_'. If specified and 'inherit.aes = TRUE' (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping. |
data |
The data to be displayed in this layer. There are three options: If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot. A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify for which variables will be created. A function will be called with a single argument, the plot data. The return value must be a data.frame., and will be used as the layer data. |
conf_level |
The percentile for the confidence interval (should fall between 0 and 1). The default is 0.95, which corresponds to a 95 percent confidence interval. |
distribution |
The distribution which the data follow, used for
calculating confidence intervals. The options are "normal", "lognormal",
and "binomial". The "normal" option will use the Student t Distribution
to calculate confidence intervals, the "lognormal" option will transform
data to the log space first. The "binomial" option will use the
|
bins |
number of bins to cut up the x data, cuts data into quantiles. |
breaks |
breaks to cut up the x data, if this option is used, bins is ignored |
geom |
Use to override the default geom. Can be a list of multiple geoms, e.g. list("point","line","errorbar"), which is the default. |
position |
Position adjustment, either as a string, or the result of a call to a position adjustment function. |
fun.args |
Optional additional arguments passed on to the functions. |
fun.data |
A function that is given the complete data and should return a data frame with variables ymin, y, and ymax. |
na.rm |
If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed. |
orientation |
The orientation of the layer, passed on to ggplot2::stat_summary.
Only implemented for ggplot2 v.3.3.0 and later. The default ("x") summarizes y values over
x values (same behavior as ggplot2 v.3.2.1 or earlier). Setting |
show.legend |
logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. |
inherit.aes |
If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders. |
... |
other arguments passed on to layer. These are often aesthetics, used to set an aesthetic to a fixed value, like color = "red" or size = 3. They may also be parameters to the paired geom/stat. |
This function can be used to generate mean +/- confidence interval plots for different distributions, and multiple geoms with a single function call.
ggplot2 plot layer
# default settings for normally distributed data, 95% confidence interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95) # try different geom xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, geom = list("ribbon", "point", "line")) # plotting lognormally distributed data data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = 10^(rep(c(1, 2, 3), each = 20) + stats::rnorm(60)), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "lognormal") # note: you DO NOT need to use both distribution = "lognormal" # and scale_y_log10() xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95) + xgx_scale_y_log10() # plotting binomial data data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = stats::rbinom(60, 1, rep(c(0.2, 0.6, 0.8), each = 20)), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "binomial") # including multiple groups in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "binomial", ggplot2::aes(color = factor(group)), position = ggplot2::position_dodge(width = 0.5)) # plotting ordinal or multinomial data set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Mild","Moderate","Severe"), 100, replace = TRUE), covariate = sample(c("Male","Female"), 100, replace = TRUE)) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = x, response = response, colour = covariate), distribution = "ordinal", bins = 4) + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~response) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = x, response = response, colour = response), distribution = "ordinal", bins = 4) + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) # Example plotting categorical vs categorical data set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Trt1", "Trt2", "Trt3"), 100, replace = TRUE), covariate = factor( sample(c("White","Black","Asian","Other"), 100, replace = TRUE), levels = c("White", "Black", "Asian", "Other"))) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = response, response = covariate), distribution = "ordinal") + xgx_stat_ci(mapping = ggplot2::aes(x = 1, response = covariate), geom = "hline", distribution = "ordinal") + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) + ggplot2::xlab("Treatment group") + ggplot2::ylab("Percent of subjects by category") # Same example with orientation flipped (only works for ggplot2 v.3.3.0 or later) # only run if ggplot2 v.3.3.0 or later ggplot2_geq_v3.3.0 <- utils::compareVersion( as.character(utils::packageVersion("ggplot2")), '3.3.0') >= 0 if(ggplot2_geq_v3.3.0){ xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(y = response, response = covariate), orientation = "y", distribution = "ordinal") + xgx_stat_ci(mapping = ggplot2::aes(y = 1, response = covariate), orientation = "y", geom = "vline", distribution = "ordinal") + ggplot2::scale_x_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) + ggplot2::ylab("Treatment group") + ggplot2::xlab("Percent of subjects by category") }# default settings for normally distributed data, 95% confidence interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95) # try different geom xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, geom = list("ribbon", "point", "line")) # plotting lognormally distributed data data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = 10^(rep(c(1, 2, 3), each = 20) + stats::rnorm(60)), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "lognormal") # note: you DO NOT need to use both distribution = "lognormal" # and scale_y_log10() xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95) + xgx_scale_y_log10() # plotting binomial data data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = stats::rbinom(60, 1, rep(c(0.2, 0.6, 0.8), each = 20)), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "binomial") # including multiple groups in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_ci(conf_level = 0.95, distribution = "binomial", ggplot2::aes(color = factor(group)), position = ggplot2::position_dodge(width = 0.5)) # plotting ordinal or multinomial data set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Mild","Moderate","Severe"), 100, replace = TRUE), covariate = sample(c("Male","Female"), 100, replace = TRUE)) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = x, response = response, colour = covariate), distribution = "ordinal", bins = 4) + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~response) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = x, response = response, colour = response), distribution = "ordinal", bins = 4) + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) # Example plotting categorical vs categorical data set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Trt1", "Trt2", "Trt3"), 100, replace = TRUE), covariate = factor( sample(c("White","Black","Asian","Other"), 100, replace = TRUE), levels = c("White", "Black", "Asian", "Other"))) xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(x = response, response = covariate), distribution = "ordinal") + xgx_stat_ci(mapping = ggplot2::aes(x = 1, response = covariate), geom = "hline", distribution = "ordinal") + ggplot2::scale_y_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) + ggplot2::xlab("Treatment group") + ggplot2::ylab("Percent of subjects by category") # Same example with orientation flipped (only works for ggplot2 v.3.3.0 or later) # only run if ggplot2 v.3.3.0 or later ggplot2_geq_v3.3.0 <- utils::compareVersion( as.character(utils::packageVersion("ggplot2")), '3.3.0') >= 0 if(ggplot2_geq_v3.3.0){ xgx_plot(data = data) + xgx_stat_ci(mapping = ggplot2::aes(y = response, response = covariate), orientation = "y", distribution = "ordinal") + xgx_stat_ci(mapping = ggplot2::aes(y = 1, response = covariate), orientation = "y", geom = "vline", distribution = "ordinal") + ggplot2::scale_x_continuous(labels = scales::percent_format()) + ggplot2::facet_wrap(~covariate) + ggplot2::ylab("Treatment group") + ggplot2::xlab("Percent of subjects by category") }
xgx_stat_pi returns a ggplot layer plotting median +/- percent
intervals
xgx_stat_pi( mapping = NULL, data = NULL, percent_level = 0.95, geom = list("line", "ribbon"), position = "identity", bins = NULL, breaks = NULL, fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )xgx_stat_pi( mapping = NULL, data = NULL, percent_level = 0.95, geom = list("line", "ribbon"), position = "identity", bins = NULL, breaks = NULL, fun.args = list(), na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, ... )
mapping |
Set of aesthetic mappings created by 'aes' or 'aes_'. If specified and 'inherit.aes = TRUE' (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping. |
data |
The data to be displayed in this layer. There are three options: If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot. A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify for which variables will be created. A function will be called with a single argument, the plot data. The return value must be a data.frame., and will be used as the layer data. |
percent_level |
The upper or lower percentile for the percent interval (should fall between 0 and 1). The default is 0.95, which corresponds to (0.05, 0.95) interval. Supplying 0.05 would give the same result |
geom |
Use to override the default geom. Can be a list of multiple geoms, e.g. list("line","ribbon"), which is the default. |
position |
Position adjustment, either as a string, or the result of a call to a position adjustment function. |
bins |
number of bins to cut up the x data, cuts data into quantiles. |
breaks |
breaks to cut up the x data, if this option is used, bins is ignored |
fun.args |
Optional additional arguments passed on to the functions. |
na.rm |
If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed. |
show.legend |
logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. |
inherit.aes |
If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders. |
... |
other arguments passed on to layer. These are often aesthetics, used to set an aesthetic to a fixed value, like color = "red" or size = 3. They may also be parameters to the paired geom/stat. |
ggplot2 plot layer
# default settings for normally distributed data, (5%,95%) interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95) # try different geom xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95, geom = list("errorbar", "point", "line")) # including multiple groups in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95, ggplot2::aes(color = factor(group), fill = factor(group)), position = ggplot2::position_dodge(width = 0.5)) # including multiple percent intervals in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.90) + xgx_stat_pi(percent_level = 0.80) + xgx_stat_pi(percent_level = 0.70) + xgx_stat_pi(percent_level = 0.60)# default settings for normally distributed data, (5%,95%) interval, data <- data.frame(x = rep(c(1, 2, 3), each = 20), y = rep(c(1, 2, 3), each = 20) + stats::rnorm(60), group = rep(1:3, 20)) xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95) # try different geom xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95, geom = list("errorbar", "point", "line")) # including multiple groups in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.95, ggplot2::aes(color = factor(group), fill = factor(group)), position = ggplot2::position_dodge(width = 0.5)) # including multiple percent intervals in same plot xgx_plot(data, ggplot2::aes(x = x, y = y)) + xgx_stat_pi(percent_level = 0.90) + xgx_stat_pi(percent_level = 0.80) + xgx_stat_pi(percent_level = 0.70) + xgx_stat_pi(percent_level = 0.60)
xgx_stat_smooth and xgx_geom_smooth produce smooth fits through continuous or categorical data.
For categorical, ordinal, or multinomial data use method = polr.
This wrapper also works with nonlinear methods like nls and nlsLM for continuous data.
xgx_geom_smooth_emax uses minpack.lm::nlsLM, predictdf.nls, and stat_smooth to display Emax model fit to data
xgx_stat_smooth( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = NULL, formula = NULL, se = TRUE, n = 80, span = 0.75, n_boot = 200, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE ) xgx_geom_smooth( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = NULL, formula = NULL, se = TRUE, n = 80, span = 0.75, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE ) xgx_geom_smooth_emax( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = "nlsLM", formula, se = TRUE, n = 80, span = 0.75, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE )xgx_stat_smooth( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = NULL, formula = NULL, se = TRUE, n = 80, span = 0.75, n_boot = 200, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE ) xgx_geom_smooth( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = NULL, formula = NULL, se = TRUE, n = 80, span = 0.75, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE ) xgx_geom_smooth_emax( mapping = NULL, data = NULL, geom = "smooth", position = "identity", ..., method = "nlsLM", formula, se = TRUE, n = 80, span = 0.75, fullrange = FALSE, level = 0.95, method.args = list(), na.rm = FALSE, orientation = "x", show.legend = NA, inherit.aes = TRUE )
mapping |
Set of aesthetic mappings created by 'aes' or 'aes_'. If specified and 'inherit.aes = TRUE' (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping. Warning: for 'method = polr', do not define 'y' aesthetic, use 'response' instead. |
data |
The data to be displayed in this layer. There are three options: If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot. A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify for which variables will be created. A function will be called with a single argument, the plot data. The return value must be a data.frame., and will be used as the layer data. |
geom |
Use to override the default geom. Can be a list of multiple geoms, e.g. list("point","line","errorbar"), which is the default. |
position |
Position adjustment, either as a string, or the result of a call to a position adjustment function. |
... |
other arguments passed on to layer. These are often aesthetics, used to set an aesthetic to a fixed value, like color = "red" or size = 3. They may also be parameters to the paired geom/stat. |
method |
method (function) to use, eg. lm, glm, gam, loess, rlm. Example: '"polr"' for ordinal data. '"nlsLM"' for nonlinear least squares. If method is left as 'NULL', then a typical 'StatSmooth' is applied, with the corresponding defaults, i.e. For datasets with n < 1000 default is loess. For datasets with 1000 or more observations defaults to gam. |
formula |
formula to use in smoothing function, eg. y ~ x, y ~ poly(x, 2), y ~ log(x) |
se |
display confidence interval around smooth? (TRUE by default, see level to control) |
n |
number of points to evaluate smoother at |
span |
Controls the amount of smoothing for the default loess smoother. Smaller numbers produce wigglier lines, larger numbers produce smoother lines. |
n_boot |
number of bootstraps to perform to compute confidence interval, currently only used for method = "polr", default is 200 |
fullrange |
should the fit span the full range of the plot, or just the data |
level |
The percentile for the confidence interval (should fall between 0 and 1). The default is 0.95, which corresponds to a 95 percent confidence interval. |
method.args |
Optional additional arguments passed on to the method. |
na.rm |
If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed. |
orientation |
The orientation of the layer, passed on to ggplot2::stat_summary.
Only implemented for ggplot2 v.3.3.0 and later. The default ("x") summarizes y values over
x values (same behavior as ggplot2 v.3.2.1 or earlier). Setting |
show.legend |
logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. |
inherit.aes |
If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders. |
ggplot2 plot layer
nlsLM uses nls.lm which implements the Levenberg-Marquardt
algorithm for fitting a nonlinear model, and may fail to converge for a
number of reasons. See ?nls.lm for more information.
nls uses Gauss-Newton method for estimating parameters,
and could fail if the parameters are not identifiable. If this happens
you will see the following warning message:
Warning message:
Computation failed in 'stat_smooth()':
singular gradient
nls will also fail if used on artificial "zero-residual" data,
use nlsLM instead.
predictdf.nls for information on how nls confidence intervals are calculated.
# Example with nonlinear least squares (method = "nlsLM") Nsubj <- 10 Doses <- c(0, 25, 50, 100, 200) Ntot <- Nsubj*length(Doses) times <- c(0,14,30,60,90) dat1 <- data.frame(ID = 1:(Ntot), DOSE = rep(Doses, Nsubj), PD0 = stats::rlnorm(Ntot, log(100), 1), Kout = exp(stats::rnorm(Ntot,-2, 0.3)), Imax = 1, ED50 = 25) %>% dplyr::mutate(PDSS = PD0*(1 - Imax*DOSE/(DOSE + ED50))*exp(stats::rnorm(Ntot, 0.05, 0.3))) %>% merge(data.frame(ID = rep(1:(Ntot), each = length(times)), Time = times), by = "ID") %>% dplyr::mutate(PD = ((PD0 - PDSS)*(exp(-Kout*Time)) + PDSS), PCHG = (PD - PD0)/PD0) gg <- ggplot2::ggplot(dat1 %>% subset(Time == 90), ggplot2::aes(x = DOSE, y = PCHG)) + ggplot2::geom_boxplot(ggplot2::aes(group = DOSE)) + xgx_theme() + xgx_scale_y_percentchangelog10() + ggplot2::ylab("Percent Change from Baseline") + ggplot2::xlab("Dose (mg)") gg + xgx_stat_smooth(method = "nlsLM", formula = y ~ E0 + Emax*x/(ED50 + x), method.args = list( start = list(Emax = -0.50, ED50 = 25, E0 = 0), lower = c(-Inf, 0, -Inf) ), se = TRUE) gg + xgx_geom_smooth_emax() ## Not run: # example with ordinal data (method = "polr") set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Mild","Moderate","Severe"), 100, replace = TRUE), covariate = sample(c("Male","Female"), 100, replace = TRUE)) %>% dplyr::mutate(y = (50 + 20*x/(200 + x))*exp(stats::rnorm(100, 0, 0.3))) # example coloring by the response categories xgx_plot(data = data) + xgx_stat_smooth(mapping = ggplot2::aes(x = x, response = response, colour = response, fill = response), method = "polr") + ggplot2::scale_y_continuous(labels = scales::percent_format()) # example faceting by the response categories, coloring by a different covariate xgx_plot(data = data) + xgx_stat_smooth(mapping = ggplot2::aes(x = x, response = response, colour = covariate, fill = covariate), method = "polr", level = 0.80) + ggplot2::facet_wrap(~response) + ggplot2::scale_y_continuous(labels = scales::percent_format()) ## End(Not run)# Example with nonlinear least squares (method = "nlsLM") Nsubj <- 10 Doses <- c(0, 25, 50, 100, 200) Ntot <- Nsubj*length(Doses) times <- c(0,14,30,60,90) dat1 <- data.frame(ID = 1:(Ntot), DOSE = rep(Doses, Nsubj), PD0 = stats::rlnorm(Ntot, log(100), 1), Kout = exp(stats::rnorm(Ntot,-2, 0.3)), Imax = 1, ED50 = 25) %>% dplyr::mutate(PDSS = PD0*(1 - Imax*DOSE/(DOSE + ED50))*exp(stats::rnorm(Ntot, 0.05, 0.3))) %>% merge(data.frame(ID = rep(1:(Ntot), each = length(times)), Time = times), by = "ID") %>% dplyr::mutate(PD = ((PD0 - PDSS)*(exp(-Kout*Time)) + PDSS), PCHG = (PD - PD0)/PD0) gg <- ggplot2::ggplot(dat1 %>% subset(Time == 90), ggplot2::aes(x = DOSE, y = PCHG)) + ggplot2::geom_boxplot(ggplot2::aes(group = DOSE)) + xgx_theme() + xgx_scale_y_percentchangelog10() + ggplot2::ylab("Percent Change from Baseline") + ggplot2::xlab("Dose (mg)") gg + xgx_stat_smooth(method = "nlsLM", formula = y ~ E0 + Emax*x/(ED50 + x), method.args = list( start = list(Emax = -0.50, ED50 = 25, E0 = 0), lower = c(-Inf, 0, -Inf) ), se = TRUE) gg + xgx_geom_smooth_emax() ## Not run: # example with ordinal data (method = "polr") set.seed(12345) data = data.frame(x = 120*exp(stats::rnorm(100,0,1)), response = sample(c("Mild","Moderate","Severe"), 100, replace = TRUE), covariate = sample(c("Male","Female"), 100, replace = TRUE)) %>% dplyr::mutate(y = (50 + 20*x/(200 + x))*exp(stats::rnorm(100, 0, 0.3))) # example coloring by the response categories xgx_plot(data = data) + xgx_stat_smooth(mapping = ggplot2::aes(x = x, response = response, colour = response, fill = response), method = "polr") + ggplot2::scale_y_continuous(labels = scales::percent_format()) # example faceting by the response categories, coloring by a different covariate xgx_plot(data = data) + xgx_stat_smooth(mapping = ggplot2::aes(x = x, response = response, colour = covariate, fill = covariate), method = "polr", level = 0.80) + ggplot2::facet_wrap(~response) + ggplot2::scale_y_continuous(labels = scales::percent_format()) ## End(Not run)
xgx_summarize_covariates
xgx_summarize_covariates(data, covariates = NULL, n_cts = 8)xgx_summarize_covariates(data, covariates = NULL, n_cts = 8)
data |
the dataset to check. must contain a USUBJID or ID column for subject id |
covariates |
the column names of covariates, to explore |
n_cts |
the number of unique values for a covariate to be treated as continuous, default is 8 |
list
data <- data.frame(ID = 1:10, WT0 = rnorm(10, 70, 10), SEX = round(runif(10))) x <- xgx_summarize_covariates(data, c("WT0", "SEX"))data <- data.frame(ID = 1:10, WT0 = rnorm(10, 70, 10), SEX = round(runif(10))) x <- xgx_summarize_covariates(data, c("WT0", "SEX"))
Calls xgx_check_data
xgx_summarize_data(data, covariates = NULL)xgx_summarize_data(data, covariates = NULL)
data |
the dataset to check. Must contain the above columns |
covariates |
the column names of covariates, to explore |
data.frame
covariates <- c("WEIGHTB", "SEX") check <- xgx_summarize_data(mad_missing_duplicates, covariates)covariates <- c("WEIGHTB", "SEX") check <- xgx_summarize_data(mad_missing_duplicates, covariates)
Calls the standard theme for xGx graphics
xgx_theme()xgx_theme()
xgx ggplot2 compatible theme
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10() + xgx_theme()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10() + xgx_theme()
xgx_theme_set
xgx_theme_set()xgx_theme_set()
xgx ggplot2 compatible theme
conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) xgx_theme_set() ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()conc <- 10^(seq(-3, 3, by = 0.1)) ec50 <- 1 data <- data.frame(concentration = conc, bound_receptor = 1 * conc / (conc + ec50)) xgx_theme_set() ggplot2::ggplot(data, ggplot2::aes(y = concentration, x = bound_receptor)) + ggplot2::geom_point() + ggplot2::geom_line() + xgx_scale_y_log10() + xgx_scale_x_reverselog10()