View source: R/Selbal_Functions.R
selbal | R Documentation |
Looks for a highly-associated balance with a response variable
selbal(x, y, th.imp = 0, covar = NULL, logit.acc = "AUC", logt = T,
col = c("steelblue1", "tomato1"), tab = T, draw = T,
maxV = 1e+10, zero.rep = "bayes")
x |
a |
y |
the response variable, either continuous or dichotomous. |
th.imp |
a numeric value indicating the minimum increment required in the association parameter between two consecutive steps in order to continue with the variable addition into the balance. |
covar |
|
logit.acc |
when |
logt |
|
col |
|
tab |
|
draw |
|
maxV |
|
zero.rep |
a value defining the method to use for zero - replacement.
|
opt.cri |
parameter for selecting the method to determine the optimal
number of variables. |
A list
with the following objects:
FINAL.BAL
the numeric values of the selected balance for each
sample.
POS
a vector with the variables appearing in the numerator of
the balance.
NEG
a vector with the variables appearing in the denominator of
the balance.
INC.VAR
a vector with both POS
and NUM
variables
(included variables).
ACC.Bal
a vector with the association value after each step of
the algorithm.
EVOL
a data.frame
with the variables sorted as they have
been added into the balance with the corresponding association value after
their inclusion. Only returned if tab
is TRUE
.
FINAL.P
the graphical representation of the results. Only
showed if draw = T
.
FIT.Final
the regression model taking covariates and the final
balance as the explanatory variables and y
as the response variable.
# Load data set
load("HIV.rda")
# Define x and y
x <- HIV[,1:60]
y <- HIV[,62]
# Run the algorithm
Bal <- selbal(x,y)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.