| Type: | Package |
| Title: | "Risk Model Regression and Analysis with Complex Non-Linear Models" |
| Version: | 1.4.3 |
| URL: | https://ericgiunta.github.io/Colossus/, https://github.com/ericgiunta/Colossus |
| BugReports: | https://github.com/ericgiunta/Colossus/issues |
| Description: | Performs survival analysis using general non-linear models. Risk models can be the sum or product of terms. Each term is the product of exponential/linear functions of covariates. Additionally sub-terms can be defined as a sum of exponential, linear threshold, and step functions. Cox Proportional hazards, Poisson, and Fine-Gray competing risks regression are supported. This work was sponsored by NASA Grants 80NSSC19M0161 and 80NSSC23M0129 through a subcontract from the National Council on Radiation Protection and Measurements (NCRP). The computing for this project was performed on the Beocat Research Cluster at Kansas State University, which is funded in part by NSF grants CNS-1006860, EPS-1006860, EPS-0919443, ACI-1440548, CHE-1726332, and NIH P20GM113109. |
| License: | GPL (≥ 3) |
| Imports: | Rcpp, data.table, parallel, stats, utils, rlang, methods, callr, stringr, processx, dplyr, tibble, lubridate |
| LinkingTo: | Rcpp, RcppEigen, testthat |
| SystemRequirements: | make, C++17 |
| RoxygenNote: | 7.3.3 |
| Encoding: | UTF-8 |
| Suggests: | knitr, rmarkdown, testthat, xml2, ggplot2, pandoc, spelling, survival, splines, dtplyr |
| Config/testthat/edition: | 3 |
| VignetteBuilder: | knitr |
| Language: | en-US |
| Biarch: | TRUE |
| NeedsCompilation: | yes |
| Packaged: | 2025-10-30 14:30:38 UTC; user |
| Author: | Eric Giunta  [aut,
cre],
Amir Bahadori
[ctb],
Dan Andresen
[ctb],
Linda Walsh [ctb],
Benjamin French
[ctb],
Lawrence Dauer
[ctb],
John Boice Jr
[ctb],
Kansas State University [cph],
NASA [fnd],
NCRP [fnd],
NRC [fnd] |
| Maintainer: | Eric Giunta <egiunta@ksu.edu> |
| Repository: | CRAN |
| Date/Publication: | 2025-10-30 15:50:02 UTC |
Fully runs a case-control regression model, returning the model and results
Description
CaseControlRun uses a formula, data.table, and list of controls to prepare and
run a Colossus matched case-control regression function
Usage
CaseControlRun(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
control = list(),
conditional_threshold = 50,
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
norm = "null",
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
conditional_threshold |
threshold above which unconditional logistic regression is used to calculate likelihoods in a matched group |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
norm |
methods used to normalize the covariates. Default is 'null' for no normalization. Other options include 'max' to normalize by the absolute maximum and 'mean' to normalize by the mean |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Case Control Wrapper Functions:
RunCaseControlRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- CaseCon_Strata(Cancer_Status, e) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- CaseControlRun(formula, df,
a_n = list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4)),
control = control
)
Automatically checks the number of starting guesses
Description
Check_Iters checks the number of iterations and number of guesses, and corrects
Usage
Check_Iters(control, a_n)
Arguments
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
Value
returns a list with the corrected control list and a_n
See Also
Other Data Cleaning Functions:
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Interprets basic cox survival formula RHS
Description
ColossusCoxSurv assigns and interprets interval columns for cox model.
This functions is called using the arguments for Cox in the right-hand side of
the formula. Uses an interval start time, end time, and event status. These are
expected to be in order or named: tstart, tend, and event.
The Fine-Gray and Stratified versions use strata and weight named options
or the last two entries.
Usage
ColossusCoxSurv(...)
Arguments
... |
entries for a cox survival object, tstart, tend, and event. Either in order or named. If unnamed and two entries, tend and event are assumed. |
Value
returns list with interval endpoints and event
See Also
Other Formula Interpretation:
ColossusLogitSurv(),
ColossusPoisSurv(),
get_form(),
get_form_joint()
Interprets basic logistic survival formula RHS with no grouping
Description
ColossusLogitSurv assigns and interprets columns for trials and events in logistic model with no grouping.
Usage
ColossusLogitSurv(...)
Arguments
... |
entries for a Logistic object, trials and events. trials not provided assumes one trial per row. |
Value
returns list with event
See Also
Other Formula Interpretation:
ColossusCoxSurv(),
ColossusPoisSurv(),
get_form(),
get_form_joint()
Interprets basic poisson survival formula RHS
Description
ColossusPoisSurv assigns and interprets interval columns for poisson model.
This functions is called using the arguments for Poisson or Poisson_Strata in the
right-hand side of the formula. Uses an person-year column, number of events, and
any strata columns. The first two are expected to be in order or named: pyr and event.
Anything beyond the event name is assumed to be strata if Poisson_Strata is used.
Usage
ColossusPoisSurv(...)
Arguments
... |
entries for a Poisson object with or without strata, pyr, event, and any strata columns. Either in order or named. The first two are assumed to be pyr and event, the rest assumed to be strata columns |
Value
returns list with duration, strata if used, and event
See Also
Other Formula Interpretation:
ColossusCoxSurv(),
ColossusLogitSurv(),
get_form(),
get_form_joint()
Fully runs a cox or fine-gray regression model, returning the model and results
Description
CoxRun uses a formula, data.table, and list of controls to prepare and
run a Colossus cox or fine-gray regression function
Usage
CoxRun(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
control = list(),
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
norm = "null",
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
norm |
methods used to normalize the covariates. Default is 'null' for no normalization. Other options include 'max' to normalize by the absolute maximum and 'mean' to normalize by the mean |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Cox Wrapper Functions:
CoxRunMulti(),
LikelihoodBound.coxres(),
RunCoxRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- Cox(Starting_Age, Ending_Age, Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- CoxRun(formula, df,
a_n = list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4)),
control = control
)
Fully runs a cox or fine-gray regression model with multiple column realizations, returning the model and results
Description
CoxRunMulti uses a formula, data.table, and list of controls to prepare and
run a Colossus cox or fine-gray regression function
Usage
CoxRunMulti(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
realization_columns = matrix(c("temp00", "temp01", "temp10", "temp11"), nrow = 2),
realization_index = c("temp0", "temp1"),
control = list(),
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
fma = FALSE,
mcml = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
realization_columns |
used for multi-realization regressions. Matrix of column names with rows for each column with realizations, columns for each realization |
realization_index |
used for multi-realization regressions. Vector of column names, one for each column with realizations. Each name should be used in the "names" variable in the equation definition |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
fma |
a boolean to denote that the Frequentist Model Averaging method should be used |
mcml |
a boolean to denote that the Monte Carlo Maximum Likelihood method should be used |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Cox Wrapper Functions:
CoxRun(),
LikelihoodBound.coxres(),
RunCoxRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"t0" = c(18, 20, 18, 19, 21, 20, 18),
"t1" = c(30, 45, 57, 47, 36, 60, 55),
"lung" = c(0, 0, 1, 0, 1, 0, 0),
"dose" = c(0, 1, 1, 0, 1, 0, 1)
)
set.seed(3742)
df$rand <- floor(runif(nrow(df), min = 0, max = 5))
df$rand0 <- floor(runif(nrow(df), min = 0, max = 5))
df$rand1 <- floor(runif(nrow(df), min = 0, max = 5))
df$rand2 <- floor(runif(nrow(df), min = 0, max = 5))
names <- c("dose", "rand")
realization_columns <- matrix(c("rand0", "rand1", "rand2"), nrow = 1)
realization_index <- c("rand")
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiter" = 1,
"halfmax" = 2, "epsilon" = 1e-6,
"deriv_epsilon" = 1e-6, "step_max" = 1.0,
"thres_step_max" = 100.0,
"verbose" = 0, "ties" = "breslow", "double_step" = 1
)
formula <- Cox(t0, t1, lung) ~ loglinear(dose, rand, 0) + multiplicative()
res <- CoxRun(formula, df, control = control)
Automates creating a date difference column
Description
Date_Shift generates a new dataframe with a column containing time difference in a given unit
Usage
Date_Shift(df, dcol0, dcol1, col_name, units = "days")
Arguments
df |
a data.table containing the columns of interest |
dcol0 |
list of starting month, day, and year |
dcol1 |
list of ending month, day, and year |
col_name |
vector of new column names |
units |
time unit to use |
Value
returns the updated dataframe
See Also
Other Data Cleaning Functions:
Check_Iters(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
m0 <- c(1, 1, 2, 2)
m1 <- c(2, 2, 3, 3)
d0 <- c(1, 2, 3, 4)
d1 <- c(6, 7, 8, 9)
y0 <- c(1990, 1991, 1997, 1998)
y1 <- c(2001, 2003, 2005, 2006)
df <- data.table::data.table("m0" = m0, "m1" = m1, "d0" = d0, "d1" = d1, "y0" = y0, "y1" = y1)
df <- Date_Shift(df, c("m0", "d0", "y0"), c("m1", "d1", "y1"), "date_since")
Generic background/excess event calculation function
Description
EventAssignment Generic background/excess event calculation function
Usage
EventAssignment(x, df, ...)
Arguments
x |
result object from a regression, class poisres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Predicts how many events are due to baseline vs excess
Description
EventAssignment Generic background/excess event calculation function, by default nothing happens
Usage
## Default S3 method:
EventAssignment(x, df, ...)
Arguments
x |
result object from a regression, class poisres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Predicts how many events are due to baseline vs excess for a completed poisson model
Description
EventAssignment.poisres uses user provided data, person-year/event columns, vectors specifying the model,
and options to calculate background and excess events for a solved Poisson regression
Usage
## S3 method for class 'poisres'
EventAssignment(
x,
df,
assign_control = list(),
control = list(),
a_n = c(),
...
)
Arguments
x |
result object from a regression, class poisres |
df |
a data.table containing the columns of interest |
assign_control |
control list for bounds calculated |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
... |
can include the named entries for the assign_control list parameter |
Value
returns a list of the final results
See Also
Other Poisson Wrapper Functions:
LikelihoodBound.poisres(),
PoisRun(),
PoisRunJoint(),
PoisRunMulti(),
Residual.poisres(),
RunPoissonRegression_Omnibus()
uses a table, list of categories, and list of event summaries to generate person-count tables
Description
Event_Count_Gen generates event-count tables
Usage
Event_Count_Gen(table, categ, events, verbose = FALSE)
Arguments
table |
dataframe with every category/event column needed |
categ |
list with category columns and methods, methods can be either strings or lists of boundaries |
events |
list of columns to summarize, supports counts and means and renaming the summary column |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns a grouped table and a list of category boundaries used
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
a <- c(0, 1, 2, 3, 4, 5, 6)
b <- c(1, 2, 3, 4, 5, 6, 7)
c <- c(0, 1, 0, 0, 0, 1, 0)
table <- data.table::data.table(
"a" = a,
"b" = b,
"c" = c
)
categ <- list(
"a" = "0/3/5]7",
"b" = list(
lower = c(-1, 3, 6),
upper = c(3, 6, 10),
name = c("low", "medium", "high")
)
)
event <- list(
"c" = "count AS cases",
"a" = "mean", "b" = "mean"
)
e <- Event_Count_Gen(table, categ, event, T)
uses a table, list of categories, list of summaries, list of events, and person-year information to generate person-time tables
Description
Event_Time_Gen generates event-time tables
Usage
Event_Time_Gen(table, pyr, categ, summaries, events, verbose = FALSE)
Arguments
table |
dataframe with every category/event column needed |
pyr |
list with entry and exit lists, containing day/month/year columns in the table |
categ |
list with category columns and methods, methods can be either strings or lists of boundaries, includes a time category or entry/exit are both required for the pyr list |
summaries |
list of columns to summarize, supports counts, means, and weighted means by person-year and renaming the summary column |
events |
list of events or interests, checks if events are within each time interval |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns a grouped table and a list of category boundaries used
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
a <- c(0, 1, 2, 3, 4, 5, 6)
b <- c(1, 2, 3, 4, 5, 6, 7)
c <- c(0, 1, 0, 0, 0, 1, 0)
d <- c(1, 2, 3, 4, 5, 6, 7)
e <- c(2, 3, 4, 5, 6, 7, 8)
f <- c(
1900, 1900, 1900, 1900,
1900, 1900, 1900
)
g <- c(1, 2, 3, 4, 5, 6, 7)
h <- c(2, 3, 4, 5, 6, 7, 8)
i <- c(
1901, 1902, 1903, 1904,
1905, 1906, 1907
)
table <- data.table::data.table(
"a" = a, "b" = b, "c" = c,
"d" = d, "e" = e, "f" = f,
"g" = g, "h" = h, "i" = i
)
categ <- list(
"a" = "-1/3/5]7",
"b" = list(
lower = c(-1, 3, 6), upper = c(3, 6, 10),
name = c("low", "medium", "high")
),
"time AS time" = list(
"day" = c(1, 1, 1, 1, 1),
"month" = c(1, 1, 1, 1, 1),
"year" = c(1899, 1903, 1910)
)
)
summary <- list(
"c" = "count AS cases",
"a" = "mean",
"b" = "weighted_mean"
)
events <- list("c")
pyr <- list(
entry = list(year = "f", month = "e", day = "d"),
exit = list(year = "i", month = "h", day = "g"),
unit = "years"
)
e <- Event_Time_Gen(table, pyr, categ, summary, events, T)
Automates creating data for a joint competing risks analysis
Description
Joint_Multiple_Events generates input for a regression with multiple non-independent events and models
Usage
Joint_Multiple_Events(
df,
events,
name_list,
term_n_list = list(),
tform_list = list(),
keep_constant_list = list(),
a_n_list = list()
)
Arguments
df |
a data.table containing the columns of interest |
events |
vector of event column names |
name_list |
list of vectors for columns for event specific or shared model elements, required |
term_n_list |
list of vectors for term numbers for event specific or shared model elements, defaults to term 0 |
tform_list |
list of vectors for subterm types for event specific or shared model elements, defaults to loglinear |
keep_constant_list |
list of vectors for constant elements for event specific or shared model elements, defaults to free (0) |
a_n_list |
list of vectors for parameter values for event specific or shared model elements, defaults to term 0 |
Value
returns the updated dataframe and model inputs
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
a <- c(0, 0, 0, 1, 1, 1)
b <- c(1, 1, 1, 2, 2, 2)
c <- c(0, 1, 2, 2, 1, 0)
d <- c(1, 1, 0, 0, 1, 1)
e <- c(0, 1, 1, 1, 0, 0)
df <- data.table("t0" = a, "t1" = b, "e0" = c, "e1" = d, "fac" = e)
time1 <- "t0"
time2 <- "t1"
df$pyr <- df$t1 - df$t0
pyr <- "pyr"
events <- c("e0", "e1")
names_e0 <- c("fac")
names_e1 <- c("fac")
names_shared <- c("t0", "t0")
term_n_e0 <- c(0)
term_n_e1 <- c(0)
term_n_shared <- c(0, 0)
tform_e0 <- c("loglin")
tform_e1 <- c("loglin")
tform_shared <- c("quad_slope", "loglin_top")
keep_constant_e0 <- c(0)
keep_constant_e1 <- c(0)
keep_constant_shared <- c(0, 0)
a_n_e0 <- c(-0.1)
a_n_e1 <- c(0.1)
a_n_shared <- c(0.001, -0.02)
name_list <- list("shared" = names_shared, "e0" = names_e0, "e1" = names_e1)
term_n_list <- list("shared" = term_n_shared, "e0" = term_n_e0, "e1" = term_n_e1)
tform_list <- list("shared" = tform_shared, "e0" = tform_e0, "e1" = tform_e1)
keep_constant_list <- list(
"shared" = keep_constant_shared,
"e0" = keep_constant_e0, "e1" = keep_constant_e1
)
a_n_list <- list("shared" = a_n_shared, "e0" = a_n_e0, "e1" = a_n_e1)
val <- Joint_Multiple_Events(
df, events, name_list, term_n_list,
tform_list, keep_constant_list, a_n_list
)
Generic likelihood boundary calculation function
Description
LikelihoodBound Generic likelihood boundary calculation function
Usage
LikelihoodBound(x, df, curve_control = list(), control = list(), ...)
Arguments
x |
result object from a regression, class coxres or poisres |
df |
a data.table containing the columns of interest |
curve_control |
a list of control options for the likelihood boundary regression. See the Control_Options vignette for details. |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
... |
extended for other necessary parameters |
Calculates the likelihood boundary for a completed cox model
Description
LikelihoodBound.coxres solves the confidence interval for a cox model, starting at the optimum point and
iteratively optimizing end-points of intervals.
Usage
## S3 method for class 'coxres'
LikelihoodBound(x, df, curve_control = list(), control = list(), ...)
Arguments
x |
result object from a regression, class coxres |
df |
a data.table containing the columns of interest |
curve_control |
a list of control options for the likelihood boundary regression. See the Control_Options vignette for details. |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
... |
can include the named entries for the curve_control list parameter |
Value
returns a list of the final results
See Also
Other Cox Wrapper Functions:
CoxRun(),
CoxRunMulti(),
RunCoxRegression_Omnibus()
Generic likelihood boundary calculation function, default option
Description
LikelihoodBound Generic likelihood boundary calculation function, by default nothing happens
Usage
## Default S3 method:
LikelihoodBound(x, df, curve_control = list(), control = list(), ...)
Arguments
x |
result object from a regression, class coxres or poisres |
df |
a data.table containing the columns of interest |
curve_control |
a list of control options for the likelihood boundary regression. See the Control_Options vignette for details. |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
... |
extended for other necessary parameters |
Calculates the likelihood boundary for a completed Poisson model
Description
LikelihoodBound.poisres solves the confidence interval for a Poisson model, starting at the optimum point and
iteratively optimizing end-points of intervals.
Usage
## S3 method for class 'poisres'
LikelihoodBound(x, df, curve_control = list(), control = list(), ...)
Arguments
x |
result object from a regression, class poisres |
df |
a data.table containing the columns of interest |
curve_control |
a list of control options for the likelihood boundary regression. See the Control_Options vignette for details. |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
... |
can include the named entries for the curve_control list parameter |
Value
returns a list of the final results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
PoisRun(),
PoisRunJoint(),
PoisRunMulti(),
Residual.poisres(),
RunPoissonRegression_Omnibus()
Defines the likelihood ratio test
Description
Likelihood_Ratio_Test uses two models and calculates the ratio
Usage
Likelihood_Ratio_Test(alternative_model, null_model)
Arguments
alternative_model |
the new model of interest in list form, output from a Poisson regression |
null_model |
a model to compare against, in list form |
Value
returns the score statistic
Examples
library(data.table)
# In an actual example, one would run two seperate RunCoxRegression regressions,
# assigning the results to e0 and e1
e0 <- list("name" = "First Model", "LogLik" = -120)
e1 <- list("name" = "New Model", "LogLik" = -100)
score <- Likelihood_Ratio_Test(e1, e0)
Calculates Full Parameter list for Special Dose Formula
Description
Linked_Dose_Formula Calculates all parameters for linear-quadratic and linear-exponential linked formulas
Usage
Linked_Dose_Formula(tforms, paras, verbose = 0)
Arguments
tforms |
list of formula types |
paras |
list of formula parameters |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns list of full parameters
Examples
library(data.table)
tforms <- list("cov_0" = "quad", "cov_1" = "exp")
paras <- list("cov_0" = c(1, 3.45), "cov_1" = c(1.2, 4.5, 0.1))
full_paras <- Linked_Dose_Formula(tforms, paras)
Calculates The Additional Parameter For a linear-exponential formula with known maximum
Description
Linked_Lin_Exp_Para Calculates what the additional parameter would be for a desired maximum
Usage
Linked_Lin_Exp_Para(y, a0, a1_goal, verbose = 0)
Arguments
y |
point formula switch |
a0 |
linear slope |
a1_goal |
exponential maximum desired |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns parameter used by Colossus
Examples
library(data.table)
y <- 7.6
a0 <- 1.2
a1_goal <- 15
full_paras <- Linked_Lin_Exp_Para(y, a0, a1_goal)
Fully runs a logistic regression model, returning the model and results
Description
LogisticRun uses a formula, data.table, and list of controls to prepare and
run a Colossus logistic regression function
Usage
LogisticRun(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
control = list(),
gradient_control = list(),
link = "odds",
single = FALSE,
observed_info = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
norm = "null",
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
link |
Used in logistic regression, the linking function relating the input model and event probability. Current options are "odds", "ident", and "loglink" for the odds ratio, identity, and complimentary loglink options. |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
norm |
methods used to normalize the covariates. Default is 'null' for no normalization. Other options include 'max' to normalize by the absolute maximum and 'mean' to normalize by the mean |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Logistic Wrapper Functions:
RunLogisticRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- logit(Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- LogisticRun(formula, df, a_n = c(1.1, -0.1, 0.2, 0.5), control = control)
Checks the OMP flag
Description
OMP_Check Called directly from R, checks the omp flag and returns true if omp is enabled
Usage
OMP_Check()
Value
boolean: True for OMP allowed
Fully runs a poisson regression model, returning the model and results
Description
PoisRun uses a formula, data.table, and list of controls to prepare and
run a Colossus poisson regression function
Usage
PoisRun(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
control = list(),
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
norm = "null",
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
norm |
methods used to normalize the covariates. Default is 'null' for no normalization. Other options include 'max' to normalize by the absolute maximum and 'mean' to normalize by the mean |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
LikelihoodBound.poisres(),
PoisRunJoint(),
PoisRunMulti(),
Residual.poisres(),
RunPoissonRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- Pois(Ending_Age, Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- PoisRun(formula, df, a_n = c(1.1, -0.1, 0.2, 0.5), control = control)
Fully runs a joint poisson regression model, returning the model and results
Description
PoisRunJoint uses a list of formula, data.table, and list of controls to prepare and
run a Colossus poisson regression function on a joint dataset
Usage
PoisRunJoint(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
control = list(),
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
norm = "null",
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
norm |
methods used to normalize the covariates. Default is 'null' for no normalization. Other options include 'max' to normalize by the absolute maximum and 'mean' to normalize by the mean |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
LikelihoodBound.poisres(),
PoisRun(),
PoisRunMulti(),
Residual.poisres(),
RunPoissonRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"Flu_Status" = c(0, 1, 0, 0, 1, 0, 1),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula_list <- list(Pois(Ending_Age, Cancer_Status) ~ plinear(d, 0),
Pois(Ending_Age, Flu_Status) ~ loglinear(d, 0),
"shared" = Pois(Ending_Age) ~ loglinear(a, b, c, 0)
)
res <- PoisRunJoint(formula_list, df, control = control)
Fully runs a poisson regression model with multiple column realizations, returning the model and results
Description
PoisRunMulti uses a formula, data.table, and list of controls to prepare and
run a Colossus poisson regression function
Usage
PoisRunMulti(
model,
df,
a_n = list(c(0)),
keep_constant = c(0),
realization_columns = matrix(c("temp00", "temp01", "temp10", "temp11"), nrow = 2),
realization_index = c("temp0", "temp1"),
control = list(),
gradient_control = list(),
single = FALSE,
observed_info = FALSE,
fma = FALSE,
mcml = FALSE,
cons_mat = as.matrix(c(0)),
cons_vec = c(0),
...
)
Arguments
model |
either a formula written for the get_form function, or the model result from the get_form function. |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
keep_constant |
binary values to denote which parameters to change |
realization_columns |
used for multi-realization regressions. Matrix of column names with rows for each column with realizations, columns for each realization |
realization_index |
used for multi-realization regressions. Vector of column names, one for each column with realizations. Each name should be used in the "names" variable in the equation definition |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
gradient_control |
a list of control options for the gradient descent algorithm. If any value is given, a gradient descent algorithm is used instead of Newton-Raphson. See the Control_Options vignette for details |
single |
a boolean to denote that only the log-likelihood should be calculated and returned, no derivatives or iterations |
observed_info |
a boolean to denote that the observed information matrix should be used to calculate the standard error for parameters, not the expected information matrix |
fma |
a boolean to denote that the Frequentist Model Averaging method should be used |
mcml |
a boolean to denote that the Monte Carlo Maximum Likelihood method should be used |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
... |
can include the named entries for the control list parameter |
Value
returns a class fully describing the model and the regression results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
LikelihoodBound.poisres(),
PoisRun(),
PoisRunJoint(),
Residual.poisres(),
RunPoissonRegression_Omnibus()
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"t0" = c(18, 20, 18, 19, 21, 20, 18),
"t1" = c(30, 45, 57, 47, 36, 60, 55),
"lung" = c(0, 0, 1, 0, 1, 0, 0),
"dose" = c(0, 1, 1, 0, 1, 0, 1)
)
set.seed(3742)
df$rand <- floor(runif(nrow(df), min = 0, max = 5))
df$rand0 <- floor(runif(nrow(df), min = 0, max = 5))
df$rand1 <- floor(runif(nrow(df), min = 0, max = 5))
df$rand2 <- floor(runif(nrow(df), min = 0, max = 5))
names <- c("dose", "rand")
realization_columns <- matrix(c("rand0", "rand1", "rand2"), nrow = 1)
realization_index <- c("rand")
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiter" = 1,
"halfmax" = 2, "epsilon" = 1e-6,
"deriv_epsilon" = 1e-6, "step_max" = 1.0,
"thres_step_max" = 100.0,
"verbose" = 0, "ties" = "breslow", "double_step" = 1
)
formula <- Pois(t1, lung) ~ loglinear(CONST, dose, rand, 0) + multiplicative()
res <- PoisRun(formula, df, control = control)
Generic relative risk calculation function
Description
RelativeRisk Generic relative risk calculation function
Usage
RelativeRisk(x, df, ...)
Arguments
x |
result object from a regression, class coxres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Calculates hazard ratios for a reference vector
Description
coxres.RelativeRisk uses a cox result object and data, to evaluate
relative risk in the data using the risk model from the result
Usage
## S3 method for class 'coxres'
RelativeRisk(x, df, a_n = c(), ...)
Arguments
x |
result object from a regression, class coxres |
df |
a data.table containing the columns of interest |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
... |
extended to match any future parameters needed |
Value
returns a class fully describing the model and the regression results
Examples
library(data.table)
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- Cox(Starting_Age, Ending_Age, Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- CoxRun(formula, df,
a_n = list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4)),
control = control
)
RelativeRisk(res, df)
Generic relative risk calculation function, default option
Description
RelativeRisk.default Generic relative risk calculation function, by default nothing happens
Usage
## Default S3 method:
RelativeRisk(x, df, ...)
Arguments
x |
result object from a regression, class coxres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Automatically assigns missing values in listed columns
Description
Replace_Missing checks each column and fills in NA values
Usage
Replace_Missing(df, name_list, msv, verbose = FALSE)
Arguments
df |
a data.table containing the columns of interest |
name_list |
vector of string column names to check |
msv |
value to replace na with, same used for every column used |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns a filled datatable
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Time_Since(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, NA, 47, 36, NA, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0)
)
df <- Replace_Missing(df, c("Starting_Age", "Ending_Age"), 70)
Generic Residual calculation function
Description
Residual Generic Residual calculation function
Usage
Residual(x, df, ...)
Arguments
x |
result object from a regression, class coxres or poisres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Generic Residual calculation function, default option
Description
Residual.default Generic Residual calculation function, by default nothing happens
Usage
## Default S3 method:
Residual(x, df, ...)
Arguments
x |
result object from a regression, class coxres or poisres |
df |
a data.table containing the columns of interest |
... |
extended for other necessary parameters |
Calculates the Residuals for a completed poisson model
Description
Residual.poisres uses user provided data, person-year/event columns, vectors specifying the model,
and options to calculate residuals for a solved Poisson regression
Usage
## S3 method for class 'poisres'
Residual(
x,
df,
control = list(),
a_n = c(),
pearson = FALSE,
deviance = FALSE,
...
)
Arguments
x |
result object from a regression, class poisres |
df |
a data.table containing the columns of interest |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
pearson |
boolean to calculate pearson residuals |
deviance |
boolean to calculate deviance residuals |
... |
can include the named entries for the assign_control list parameter |
Value
returns a list of the final results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
LikelihoodBound.poisres(),
PoisRun(),
PoisRunJoint(),
PoisRunMulti(),
RunPoissonRegression_Omnibus()
Performs Matched Case-Control Conditional Logistic Regression
Description
RunCaseControlRegression_Omnibus uses user provided data, time/event columns,
vectors specifying the model, and options to control the convergence
and starting positions. Has additional options for starting with several
initial guesses, using stratification and/or matching by time at risk,
and calculation without derivatives
Usage
RunCaseControlRegression_Omnibus(
df,
time1 = "%trunc%",
time2 = "%trunc%",
event0 = "event",
names = c("CONST"),
term_n = c(0),
tform = "loglin",
keep_constant = c(0),
a_n = c(0),
modelform = "M",
control = list(),
strat_col = "null",
cens_weight = "null",
model_control = list(),
cons_mat = as.matrix(c(0)),
cons_vec = c(0)
)
Arguments
df |
a data.table containing the columns of interest |
time1 |
column used for time period starts |
time2 |
column used for time period end |
event0 |
column used for event status |
names |
columns for elements of the model, used to identify data columns |
term_n |
term numbers for each element of the model |
tform |
list of string function identifiers, used for linear/step |
keep_constant |
binary values to denote which parameters to change |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
modelform |
string specifying the model type: M, ME, A, PA, PAE, GMIX, GMIX-R, GMIX-E |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
strat_col |
column to stratify by if needed |
cens_weight |
column containing the row weights |
model_control |
controls which alternative model options are used, see the Control_Options vignette for further details |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
Value
returns a list of the final results
See Also
Other Case Control Wrapper Functions:
CaseControlRun()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
# For the interval case
time1 <- "Starting_Age"
time2 <- "Ending_Age"
event <- "Cancer_Status"
names <- c("a", "b", "c", "d")
a_n <- list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4))
# used to test at a specific point
term_n <- c(0, 1, 1, 2)
tform <- c("loglin", "lin", "lin", "plin")
modelform <- "M"
keep_constant <- c(0, 0, 0, 0)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(5, 5, 5),
"halfmax" = 5, "epsilon" = 1e-3, "deriv_epsilon" = 1e-3,
"step_max" = 1.0, "thres_step_max" = 100.0,
"verbose" = FALSE,
"ties" = "breslow", "double_step" = 1
)
e <- RunCaseControlRegression_Omnibus(df, time1, time2, event,
names, term_n, tform, keep_constant,
a_n, modelform, control,
model_control = list(
"stata" = FALSE,
"time_risk" = FALSE
)
)
Performs Cox Proportional Hazards regression using the omnibus function
Description
RunCoxRegression_Omnibus uses user provided data, time/event columns,
vectors specifying the model, and options to control the convergence
and starting positions. Has additional options for starting with several
initial guesses, using stratification, multiplicative loglinear 1-term,
competing risks, and calculation without derivatives
Usage
RunCoxRegression_Omnibus(
df,
time1 = "%trunc%",
time2 = "%trunc%",
event0 = "event",
names = c("CONST"),
term_n = c(0),
tform = "loglin",
keep_constant = c(0),
a_n = c(0),
modelform = "M",
control = list(),
strat_col = "null",
cens_weight = "null",
model_control = list(),
cons_mat = as.matrix(c(0)),
cons_vec = c(0)
)
Arguments
df |
a data.table containing the columns of interest |
time1 |
column used for time period starts |
time2 |
column used for time period end |
event0 |
column used for event status |
names |
columns for elements of the model, used to identify data columns |
term_n |
term numbers for each element of the model |
tform |
list of string function identifiers, used for linear/step |
keep_constant |
binary values to denote which parameters to change |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
modelform |
string specifying the model type: M, ME, A, PA, PAE, GMIX, GMIX-R, GMIX-E |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
strat_col |
column to stratify by if needed |
cens_weight |
column containing the row weights |
model_control |
controls which alternative model options are used, see the Control_Options vignette for further details |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
Value
returns a list of the final results
See Also
Other Cox Wrapper Functions:
CoxRun(),
CoxRunMulti(),
LikelihoodBound.coxres()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
# For the interval case
time1 <- "Starting_Age"
time2 <- "Ending_Age"
event <- "Cancer_Status"
names <- c("a", "b", "c", "d")
a_n <- list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4))
# used to test at a specific point
term_n <- c(0, 1, 1, 2)
tform <- c("loglin", "lin", "lin", "plin")
modelform <- "M"
keep_constant <- c(0, 0, 0, 0)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(5, 5, 5),
"halfmax" = 5, "epsilon" = 1e-3, "deriv_epsilon" = 1e-3,
"step_max" = 1.0, "thres_step_max" = 100.0,
"verbose" = FALSE,
"ties" = "breslow", "double_step" = 1, "guesses" = 2
)
e <- RunCoxRegression_Omnibus(df, time1, time2, event,
names, term_n, tform, keep_constant,
a_n, modelform, control,
model_control = list(
"single" = FALSE,
"basic" = FALSE, "cr" = FALSE, "null" = FALSE
)
)
Performs basic Logistic regression using the omnibus function
Description
RunLogisticRegression_Omnibus uses user provided data, time/event columns,
vectors specifying the model, and options to control the convergence and starting positions.
Has additional options to starting with several initial guesses
Usage
RunLogisticRegression_Omnibus(
df,
trial0 = "CONST",
event0 = "event",
names = c("CONST"),
term_n = c(0),
tform = "loglin",
keep_constant = c(0),
a_n = c(0),
modelform = "M",
control = list(),
model_control = list(),
cons_mat = as.matrix(c(0)),
cons_vec = c(0)
)
Arguments
df |
a data.table containing the columns of interest |
trial0 |
column with the number of trials per row, assumed to be 1 if a column not provided |
event0 |
column used for event status |
names |
columns for elements of the model, used to identify data columns |
term_n |
term numbers for each element of the model |
tform |
list of string function identifiers, used for linear/step |
keep_constant |
binary values to denote which parameters to change |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
modelform |
string specifying the model type: M, ME, A, PA, PAE, GMIX, GMIX-R, GMIX-E |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
model_control |
controls which alternative model options are used, see the Control_Options vignette for further details |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
Value
returns a list of the final results
See Also
Other Logistic Wrapper Functions:
LogisticRun()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"Trials" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
# For the interval case
trial <- "Trials"
event <- "Cancer_Status"
names <- c("a", "b", "c", "d")
a_n <- c(1.1, -0.1, 0.2, 0.5) # used to test at a specific point
term_n <- c(0, 1, 1, 2)
tform <- c("loglin", "lin", "lin", "plin")
modelform <- "M"
keep_constant <- c(0, 0, 0, 0)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiter" = 5,
"halfmax" = 5, "epsilon" = 1e-3,
"deriv_epsilon" = 1e-3, "step_max" = 1.0,
"thres_step_max" = 100.0, "verbose" = FALSE, "ties" = "breslow",
"double_step" = 1
)
strat_col <- "e"
e <- RunLogisticRegression_Omnibus(
df, trial, event, names, term_n,
tform, keep_constant,
a_n, modelform,
control
)
Performs basic Poisson regression using the omnibus function
Description
RunPoissonRegression_Omnibus uses user provided data, time/event columns,
vectors specifying the model, and options to control the convergence and starting positions.
Has additional options to starting with several initial guesses
Usage
RunPoissonRegression_Omnibus(
df,
pyr0 = "pyr",
event0 = "event",
names = c("CONST"),
term_n = c(0),
tform = "loglin",
keep_constant = c(0),
a_n = c(0),
modelform = "M",
control = list(),
strat_col = "null",
model_control = list(),
cons_mat = as.matrix(c(0)),
cons_vec = c(0)
)
Arguments
df |
a data.table containing the columns of interest |
pyr0 |
column used for person-years per row |
event0 |
column used for event status |
names |
columns for elements of the model, used to identify data columns |
term_n |
term numbers for each element of the model |
tform |
list of string function identifiers, used for linear/step |
keep_constant |
binary values to denote which parameters to change |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
modelform |
string specifying the model type: M, ME, A, PA, PAE, GMIX, GMIX-R, GMIX-E |
control |
list of parameters controlling the convergence, see the Control_Options vignette for details |
strat_col |
column to stratify by if needed |
model_control |
controls which alternative model options are used, see the Control_Options vignette for further details |
cons_mat |
Matrix containing coefficients for a system of linear constraints, formatted as matrix |
cons_vec |
Vector containing constants for a system of linear constraints, formatted as vector |
Value
returns a list of the final results
See Also
Other Poisson Wrapper Functions:
EventAssignment.poisres(),
LikelihoodBound.poisres(),
PoisRun(),
PoisRunJoint(),
PoisRunMulti(),
Residual.poisres()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
# For the interval case
pyr <- "Ending_Age"
event <- "Cancer_Status"
names <- c("a", "b", "c", "d")
a_n <- c(1.1, -0.1, 0.2, 0.5) # used to test at a specific point
term_n <- c(0, 1, 1, 2)
tform <- c("loglin", "lin", "lin", "plin")
modelform <- "M"
keep_constant <- c(0, 0, 0, 0)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiter" = 5,
"halfmax" = 5, "epsilon" = 1e-3,
"deriv_epsilon" = 1e-3, "step_max" = 1.0,
"thres_step_max" = 100.0, "verbose" = FALSE, "ties" = "breslow",
"double_step" = 1
)
strat_col <- "e"
e <- RunPoissonRegression_Omnibus(
df, pyr, event, names, term_n,
tform, keep_constant,
a_n, modelform,
control, strat_col
)
Checks OS, compilers, and OMP
Description
System_Version checks OS, default R c++ compiler, and if OMP is enabled
Usage
System_Version()
Value
returns a list of results
See Also
Other Output and Information Functions:
print.caseconres(),
print.coxres(),
print.coxresbound(),
print.logitres(),
print.poisres(),
print.poisresbound()
Automates creating a date since a reference column
Description
Time_Since generates a new dataframe with a column containing time since a reference in a given unit
Usage
Time_Since(df, dcol0, tref, col_name, units = "days")
Arguments
df |
a data.table containing the columns of interest |
dcol0 |
list of ending month, day, and year |
tref |
reference time in date format |
col_name |
vector of new column names |
units |
time unit to use |
Value
returns the updated dataframe
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
apply_norm(),
factorize(),
gen_time_dep()
Examples
library(data.table)
m0 <- c(1, 1, 2, 2)
m1 <- c(2, 2, 3, 3)
d0 <- c(1, 2, 3, 4)
d1 <- c(6, 7, 8, 9)
y0 <- c(1990, 1991, 1997, 1998)
y1 <- c(2001, 2003, 2005, 2006)
df <- data.table::data.table(
"m0" = m0, "m1" = m1,
"d0" = d0, "d1" = d1,
"y0" = y0, "y1" = y1
)
tref <- strptime("3-22-1997", format = "%m-%d-%Y", tz = "UTC")
df <- Time_Since(df, c("m1", "d1", "y1"), tref, "date_since")
Automatically applies a normalization to either an input or output
Description
apply_norm applies a normalization factor
Usage
apply_norm(df, norm, names, input, values, model_control)
Arguments
df |
The data.table with columns to be normalized |
norm |
The normalization option used, currently max or mean |
names |
columns for elements of the model, used to identify data columns |
input |
boolean if the normalization is being performed on the input values or on an output |
values |
list of values using during normalization |
model_control |
controls which alternative model options are used, see the Control_Options vignette for further details |
Value
returns list with the normalized values
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
factorize(),
gen_time_dep()
Splits a parameter into factors
Description
factorize uses user provided list of columns to define new parameter for each unique value and update the data.table.
Not for interaction terms
Usage
factorize(df, col_list, verbose = 0)
Arguments
df |
a data.table containing the columns of interest |
col_list |
an array of column names that should have factor terms defined |
verbose |
integer valued 0-4 controlling what information is printed to the terminal. Each level includes the lower levels. 0: silent, 1: errors printed, 2: warnings printed, 3: notes printed, 4: debug information printed. Errors are situations that stop the regression, warnings are situations that assume default values that the user might not have intended, notes provide information on regression progress, and debug prints out C++ progress and intermediate results. The default level is 2 and True/False is converted to 3/0. |
Value
returns a list with two named fields. df for the updated dataframe, and cols for the new column names
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
gen_time_dep()
Examples
library(data.table)
a <- c(0, 1, 2, 3, 4, 5, 6)
b <- c(1, 2, 3, 4, 5, 6, 7)
c <- c(0, 1, 2, 1, 0, 1, 0)
df <- data.table::data.table("a" = a, "b" = b, "c" = c)
col_list <- c("c")
val <- factorize(df, col_list)
df <- val$df
new_col <- val$cols
Applies time dependence to parameters
Description
gen_time_dep generates a new dataframe with time dependent covariates by applying a grid in time
Usage
gen_time_dep(
df,
time1,
time2,
event0,
iscox,
dt,
new_names,
dep_cols,
func_form,
fname,
tform,
nthreads = as.numeric(detectCores())
)
Arguments
df |
a data.table containing the columns of interest |
time1 |
column used for time period starts |
time2 |
column used for time period end |
event0 |
column used for event status |
iscox |
boolean if rows not at event times should not be kept, rows are removed if true. a Cox proportional hazards model does not use rows with intervals not containing event times |
dt |
spacing in time for new rows |
new_names |
list of new names to use instead of default, default used if entry is ”" |
dep_cols |
columns that are not needed in the new dataframe |
func_form |
vector of functions to apply to each time-dependent covariate. Of the form func(df, time) returning a vector of the new column value |
fname |
filename used for new dataframe |
tform |
list of string function identifiers, used for linear/step |
nthreads |
number of threads to use, do not use more threads than available on your machine |
Value
returns the updated dataframe
See Also
Other Data Cleaning Functions:
Check_Iters(),
Date_Shift(),
Event_Count_Gen(),
Event_Time_Gen(),
Joint_Multiple_Events(),
Replace_Missing(),
Time_Since(),
apply_norm(),
factorize()
Examples
library(data.table)
# Adapted from the tests
a <- c(20, 20, 5, 10, 15)
b <- c(1, 2, 1, 1, 2)
c <- c(0, 0, 1, 1, 1)
df <- data.table::data.table("a" = a, "b" = b, "c" = c)
time1 <- "%trunc%"
time2 <- "a"
event <- "c"
control <- list(
"lr" = 0.75, "maxiter" = -1, "halfmax" = 5, "epsilon" = 1e-9,
"deriv_epsilon" = 1e-9, "step_max" = 1.0,
"thres_step_max" = 100.0,
"verbose" = FALSE, "ties" = "breslow", "double_step" = 1
)
grt_f <- function(df, time_col) {
return((df[, "b"] * df[, get(time_col)])[[1]])
}
func_form <- c("lin")
df_new <- gen_time_dep(
df, time1, time2, event, TRUE, 0.01, c("grt"), c(),
c(grt_f), paste("test", "_new.csv", sep = ""), func_form, 2
)
file.remove("test_new.csv")
Interprets a Colossus formula and makes necessary changes to data
Description
get_form uses a formula and data.table, to fully describe the model
for a Colossus regression function.
Usage
get_form(formula, df)
Arguments
formula |
a formula object, written in Colossus notation. See the Unified Equation Representation vignette for details. |
df |
a data.table containing the columns of interest |
Value
returns a class fully describing the model and the updated data
See Also
Other Formula Interpretation:
ColossusCoxSurv(),
ColossusLogitSurv(),
ColossusPoisSurv(),
get_form_joint()
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1),
"e" = c(0, 0, 1, 0, 0, 0, 1)
)
formula <- Cox(Starting_Age, Ending_Age, Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
model <- get_form(formula, df)
Interprets a Poisson joint formula and makes necessary changes to data
Description
get_form_joint uses two event formula, a shared formula,
and data.table, to fully describe the model for a joint Poisson model.
Usage
get_form_joint(formula_list, df)
Arguments
formula_list |
a list of formula objects, each written in Colossus notation. See the Unified Equation Representation vignette for details. Each formula should include the elements specific to the specified event column. The list can include an entry named "shared" to denote shared terms. The person-year and strata columns should be the same. |
df |
a data.table containing the columns of interest |
Value
returns a class fully describing the model and the updated data
See Also
Other Formula Interpretation:
ColossusCoxSurv(),
ColossusLogitSurv(),
ColossusPoisSurv(),
get_form()
Performs Cox Proportional Hazard model plots
Description
plot.coxres uses user provided data, time/event columns,
vectors specifying the model, and options to choose and save plots
Usage
## S3 method for class 'coxres'
plot(x, df, plot_options, a_n = c(), ...)
Arguments
x |
result object from a regression, class coxres |
df |
a data.table containing the columns of interest |
plot_options |
list of parameters controlling the plot options, see RunCoxPlots() for different options |
a_n |
list of initial parameter values, used to determine the number of parameters. May be either a list of vectors or a single vector. |
... |
can include the named entries for the plot_options parameter |
Value
saves the plots in the current directory and returns the data used for plots
Examples
library(data.table)
## basic example code reproduced from the starting-description vignette
df <- data.table::data.table(
"UserID" = c(112, 114, 213, 214, 115, 116, 117),
"Starting_Age" = c(18, 20, 18, 19, 21, 20, 18),
"Ending_Age" = c(30, 45, 57, 47, 36, 60, 55),
"Cancer_Status" = c(0, 0, 1, 0, 1, 0, 0),
"a" = c(0, 1, 1, 0, 1, 0, 1),
"b" = c(1, 1.1, 2.1, 2, 0.1, 1, 0.2),
"c" = c(10, 11, 10, 11, 12, 9, 11),
"d" = c(0, 0, 0, 1, 1, 1, 1)
)
control <- list(
"ncores" = 2, "lr" = 0.75, "maxiters" = c(1, 1),
"halfmax" = 1
)
formula <- Cox(Starting_Age, Ending_Age, Cancer_Status) ~
loglinear(a, b, c, 0) + plinear(d, 0) + multiplicative()
res <- CoxRun(formula, df,
control = control,
a_n = list(c(1.1, -0.1, 0.2, 0.5), c(1.6, -0.12, 0.3, 0.4))
)
plot_options <- list(
"type" = c("surv", paste(tempfile(),
"run",
sep = ""
)), "studyid" = "UserID",
"verbose" = FALSE
)
plot(res, df, plot_options)
Prints a case-control regression output clearly
Description
print.caseconres uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'caseconres'
print(x, ...)
Arguments
x |
result object from a regression, class caseconres |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.coxres(),
print.coxresbound(),
print.logitres(),
print.poisres(),
print.poisresbound()
Prints a cox regression output clearly
Description
print.coxres uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'coxres'
print(x, ...)
Arguments
x |
result object from a regression, class coxres |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.caseconres(),
print.coxresbound(),
print.logitres(),
print.poisres(),
print.poisresbound()
Prints a cox likelihood boundary regression output clearly
Description
print.coxresbound uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'coxresbound'
print(x, ...)
Arguments
x |
result object from a regression, class coxresbound |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.caseconres(),
print.coxres(),
print.logitres(),
print.poisres(),
print.poisresbound()
Prints a logistic regression output clearly
Description
print.logitres uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'logitres'
print(x, ...)
Arguments
x |
result object from a regression, class logitres |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.caseconres(),
print.coxres(),
print.coxresbound(),
print.poisres(),
print.poisresbound()
Prints a poisson regression output clearly
Description
print.poisres uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'poisres'
print(x, ...)
Arguments
x |
result object from a regression, class poisres |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.caseconres(),
print.coxres(),
print.coxresbound(),
print.logitres(),
print.poisresbound()
Prints a poisson likelihood boundary regression output clearly
Description
print.poisresbound uses the list output from a regression, prints off a table of results and summarizes the score and convergence.
Usage
## S3 method for class 'poisresbound'
print(x, ...)
Arguments
x |
result object from a regression, class poisresbound |
... |
can include the number of digits, named digit, or an unnamed integer entry assumed to be digits |
Value
return nothing, prints the results to console
See Also
Other Output and Information Functions:
System_Version(),
print.caseconres(),
print.coxres(),
print.coxresbound(),
print.logitres(),
print.poisres()