All options are set as global variables at the beginning of installing the functions with source(zcurve3). Afterwards they can be changed like any other R object
1. Curve Type: Default Z-Values, Option Fit t-Distributions with a Fixed df
CURVE.TYPE <- “z” # Set to “t” for t-distribution
df = c() # set to the df of the t-distribution
2. Speed Control Parameters
parallel <- FALSE # Placeholder – parallel functionality not yet implemented
max_iter <- 1e6 # Max iterations for model estimation
max_iter_boot <- 1e5 # Max iterations for bootstrapped estimates
EM.criterion <- 1e-3 # Convergence threshold for EM algorithm
EM.max.iter <- 1000 # Max iterations for EM
Plot.Fitting <- FALSE # Plot fitting curve (only for Est.Method = “OF” or “EXT”)
PLOT SETTINGS
Title <- “” # Optional plot title
letter.size <- 1 # Text size in plots
letter.size.1 <- letter.size # Used for version labels in plot
y.line.factor <- 3 # Controls spacing of plot text
x.lim.min <- 0 # X-axis lower bound
x.lim.max <- 6 # X-axis upper bound
ymax <- 0.6 # Y-axis upper bound
ymin <- 0 # OUTDATED Y-axis lower bound (for label space)
Show.Histogram <- TRUE # Toggle histogram in plot
Show.Text <- TRUE # Toggle model results in plot
Show.Curve.All <- TRUE # Show predicted z-curve
Show.Curve.Sig <- FALSE # Option: show z-curve only for significant values
Show.Significance <- TRUE # Show z = critical value line
Show.KD <- FALSE # Toggle kernel density overlay (density method only)
sig.levels <- c() # Optional: mark additional p-value thresholds on plot
int.loc <- 0.5 # Plot local power intervals below x-axis (set 0 to disable)
hist.bar.width <- 0.2 # Width of histogram bars
bw.draw <- 0.10 # Smoothing for kernel density display
CONSOLE OUTPUT
Show.Iterations <- TRUE # Show iterations for slow procedures (e.g., EXT, TEST4HETEROGENEITY)
MODEL PARAMETERS
alpha <- 0.05 # Significance level
crit <- qnorm(1 – alpha / 2) # Corresponding two-sided critical z
two.sided <- TRUE # Assume two-sided z-values (use abs(z)); not yet compatible with signed z-values
Color scheme
col.curve <- “violetred3”
col.hist <- “blue3”
col.kd <- “green3”
Est.Method <- “OF” # Estimation method: “OF”, “EM”, or “EXT” # Clustered Data: “CLU-W” (weighted),”CLU-B” (bootstrap) Int.Beg <- 1.96 # Default: critical value for alpha = .05 Int.End <- 6 # End of modeling interval (z > 6 = power = 1)
ncp <- 0:6 # Component locations (z-values at which densities are centered)
components <- length(ncp) # Number of components
zsd <- 1 # SD of standard normal z-distribution
zsds = rep(zsd,components) # one SD for each component
just <- 0.8 # Cutoff for “just significant” z-values (used in optional bias test)
ZSDS.FIXED <- FALSE # Fix SD values for EXT method
NCP.FIXED <- FALSE # Fix non-central parameter(NCP) means values for EXT method
W.FIXED <- FALSE # Fix weights for EXT method
fixed.false.positives <- 0 # If > 0, constrains proportion of false positives (e.g., weight for z = 0 component)
DENSITY-BASED SETTINGS (Only used with Est.Method = “OF”)
n.bars <- 512 # Number of bars in histogram
Augment <- TRUE # Apply correction for bias at lower bound
Augment.Regression <- FALSE # Use Slope for Augmentation
Augment.Factor <- 1 # Amount of augmentation
bw.est <- 0.05 # Bandwidth for kernel density (lower = less smoothing, higher = more smoothing)
bw.aug <- .20 # Width of Augmentation interval
INPUT RESTRICTIONS
MAX.INP.Z <- Inf # Optionally restrict very large z-values (set Inf to disable)
CONFIDENCE INTERVALS / BOOTSTRAPS
boot.iter <- 0 # Number of bootstrap iterations (suggest 500+ for final models)
ERR.CI.adjust <- 0.03 # Conservative widening of confidence intervals for ERR
EDR.CI.adjust <- 0.05 # Conservative widening for EDR
CI.ALPHA <- 0.05 # CI level (default = 95%)
CI levels for Heterogeneity Test
fit.ci <- c(.01, .025, .05, .10, .17, .20, .50, .80, .83, .90, .95, .975, .99) # CI levels for model fit test
TEST4BIAS <- FALSE # Enable optional bias test
TEST4HETEROGENEITY <- 0 # Optional heterogeneity test (slow) — set number of bootstrap iterations