R/Plotting_code.R
In rettopnivek/nROUSE: Functions for the nROUSE model
#----------------------------------------------#
# R code for plotting/demonstrations of nROUSE #
#----------------------------------------------#
# Index
# Lookup - 01: nROUSE_predictions
# Lookup - 02: nROUSE_demonstration
# Lookup - 01
#' Predictions from the nROUSE model
#'
#' Generates a plot of the predicted accuracy over different
#' prime durations and types for the nROUSE model (output can
#' be generated as well).
#'
#' @param primeDurations a vector of prime durations between
#' 17 and 2000 ms.
#' @param primeTypes a character vector with up to 4 types of prime:
#' 'Target', 'Foil', 'Neither', or 'Both'.
#' @param design a vector with the duration (in ms) of the target flash,
#' the target mask, and the choice options.
#' @param opt a logical vector indicating if the predicted accuracies
#' latencies should be returned, and if a plot should be generated.
#' @param prm an optional named vector for the parameters of the
#' nROUSE model, where...
#' \itemize{
#' \item Fe = the feedback scalar.
#' \item N = The noise multiplier.
#' \item L = The constant leak current.
#' \item D = The synaptic depletion rate.
#' \item R = The replenishment rate.
#' \item I = The inhibition constant.
#' \item Th = The noise multiplier.
#' \item Ta = The temporal attention parameter.
#' \item SV = The visual integration rate.
#' \item SO = The orthographic integration rate.
#' \item SS = The semantic integration rate.
#' }
#'
#'
#' @return A list consisting of a matrix of predicted accuracies by
#' prime duration and type, and an array with the predicted perceptual
#' identification latencies for the target and foil by prime duration
#' and prime type.
#'
#' @examples
#' # Default
#' nROUSE_predictions()
#'
#' # Specific conditions
#'
#' # Prime durations
#' pd = c(50,400)
#' # Prime types
#' pt = c('Target','Foil')
#' nROUSE_predictions( primeDurations = pd, primeTypes = pt )
#'
#' # Different timing for target stimulus flash and subsequent mask
#' nROUSE_predictions( design = c( 84, 416, 500 ) )
#'
#' # Suppress plotting and return predictions
#' pred = nROUSE_predictions( opt = c( T, F ) )
#'
#' # Different parameter values (Must be named)
#' prm = c( I = .5, Ta = .8, N = .025 )
#' nROUSE_predictions( prm = prm )
#'
#' @export
nROUSE_predictions = function( primeDurations = c( 17,50,150,400,2000 ),
primeTypes = c('Target','Foil','Neither','Both'),
design = c( 50, 450, 500 ),
opt = c( F, T ),
prm = NULL ) {
# Default parameters (Rieth & Huber, 2017)
param = c( Fe = .25, N = .0302, L = .15,
D = .324, R = .022, I = .9844,
Th = .15, Ta = 1.0, SV = .0294,
SO = .0609, SS = .015 )
# Check if parameter values should be changed
if ( length( prm ) > 0 ) {
selPar = names(prm)
for (i in 1:length(selPar)) {
chk = names(param) == selPar[i]
if ( sum(chk) == 1 )
param[ chk ] = prm[i]
}
}
# Sort prime duratiosn
primeDurations = sort( primeDurations );
# Check if prime types are admissable
chk = c('Target','Foil','Neither','Both')
if ( sum(primeTypes %in% chk) < length(primeTypes) )
stop('Check character vector for prime types.')
# Matrices to store output
Accuracy = matrix( NA, length( primeDurations ),
length( primeTypes ) )
colnames( Accuracy ) = primeTypes;
rownames( Accuracy ) = as.character( primeDurations )
Latency = array( NA, dim=c( length( primeDurations ),
length( primeTypes ),
2 ) )
# Simulate nROUSE
for ( pt in 1:length( primeTypes ) ) {
for ( pd in 1:length( primeDurations ) ) {
pres = c( primeDurations[pd], design )
if (primeTypes[pt] == 'Target') prmT = c(2,0)
if (primeTypes[pt] == 'Foil') prmT = c(0,2)
if (primeTypes[pt] == 'Neither') prmT = c(0,0)
if (primeTypes[pt] == 'Both') prmT = c(1,1)
sim = simulate_nROUSE( pres, prmT, param )
Accuracy[pd,pt] = sim$Latencies[3]
Latency[pd,pt,] = sim$Latencies[1:2]
}
}
# Plot results
if (opt[2]) {
clrs = c( Target = 'orange',
Foil = 'red',
Neither = 'blue',
Both = 'purple' )
xl = c( 10, max( primeDurations ) )
plot( log(xl), c(0,1), type = 'n', xaxt = 'n',
yaxt = 'n', xlab = ' ', ylab = ' ',
bty = 'l' )
xa = c(10,100,1000,5000)
axis( 1, log( xa[ xa <= max(xl) ] ),
xa[ xa <= max(xl) ], cex.axis = 1.5 )
mtext('Prime duration (log ms)',side=1,cex=1.5,
line = 2.25)
axis(2,seq(0,1,.25),cex.axis=1.5)
mtext('Accuracy',side=2,cex=1.5,
line = 2.25)
abline( h = .5, lty = 2, lwd = 2 )
mtext('nROUSE predictions', side = 3, line = 1,
cex = 1.5 )
whchClr = c()
for ( pt in 1:length( primeTypes ) ) {
clr = clrs[ names(clrs) %in% primeTypes[pt] ]
whchClr = c( whchClr, clr )
lines( log( primeDurations ), Accuracy[,pt], col = clr,
lwd = 2 )
points( log( primeDurations ), Accuracy[,pt], bg = clr,
pch = 21, cex = 1.2 )
}
legend('bottomleft',primeTypes, fill = whchClr, bty = 'n',
cex = 1.5 )
}
if (opt[1]) return( list( Accuracy = Accuracy, Latency = Latency ) )
}
# Lookup - 02
#' Demonstration of the nROUSE model.
#'
#' Generates a plot of the activation at the highest level of the
#' neural network (i.e. semantic-lexical) for the target and foil
#' representations under different prime types and an inputted
#' prime duration.
#'
#' @param primeTypes a character vector with up to 4 types of prime:
#' 'Target', 'Foil', 'Neither', or 'Both'.
#' @param design a vector with the duration (in ms) of the target flash,
#' the target mask, and the choice options.
#' @param prm an optional named vector for the parameters of the
#' nROUSE model, where...
#' \itemize{
#' \item Fe = the feedback scalar.
#' \item N = The noise multiplier.
#' \item L = The constant leak current.
#' \item D = The synaptic depletion rate.
#' \item R = The replenishment rate.
#' \item I = The inhibition constant.
#' \item Th = The noise multiplier.
#' \item Ta = The temporal attention parameter.
#' \item SV = The visual integration rate.
#' \item SO = The orthographic integration rate.
#' \item SS = The semantic integration rate.
#' }
#'
#' @details The user will receive a prompt to input a prime duration.
#' The function will then generate a plot showing the time course
#' of activation for the target/foil representations for the
#' set of prime types. The predicted accuracies and the difference
#' distribution from which these accuracies are derived are shown
#' on the right.
#'
#' @examples
#' \dontrun{
#' # Default
#' nROUSE_demonstration()
#'
#' # Fewer prime types
#' pt = c( 'Target', 'Foil' )
#' nROUSE_demonstration(primeTypes = pt)
#'
#' # Different timing for target stimulus flash and subsequent mask
#' nROUSE_demonstration(design=c(84,416,500))
#'
#' # Different parameter values (Must be named)
#' prm = c( I = .5, Ta = .8, N = .025 )
#' nROUSE_demonstration(prm=prm)
#' }
#'
#' @export
nROUSE_demonstration = function(primeTypes = c('Target','Foil',
'Neither','Both'),
design = c( 50, 450, 500 ),
prm = NULL ) {
# Default parameters
param = c( Fe = .25, N = .0302, L = .15,
D = .324, R = .022, I = .9844,
Th = .15, Ta = 1.0, SV = .0294,
SO = .0609, SS = .015 )
# Check if parameter values should be changed
if ( length( prm ) > 0 ) {
selPar = names(prm)
for (i in 1:length(selPar)) {
chk = names(param) == selPar[i]
if ( sum(chk) == 1 )
param[ chk ] = prm[i]
}
}
cat("Prime duration? (17 - 2000)","\n")
cat("Type number (in ms) and hit 'enter'","\n")
pd = scan(n = 1)
td = design[1]
md = design[2]
cd = design[3]
presentations = c( pd, design )
pltStr = c()
npt = length( primeTypes )
pltStr = matrix( rep( 1:length( primeTypes ), each = 8 ),
length( primeTypes ), 8, byrow = T )
pltStr = pltStr + 0:(length(primeTypes)-1)
pltStr = cbind( pltStr, pltStr[,1:2]+1 )
layout( pltStr )
for (pt in 1:length(primeTypes)) {
if (primeTypes[pt] == 'Target') prmT = c(2,0)
if (primeTypes[pt] == 'Foil') prmT = c(0,2)
if (primeTypes[pt] == 'Neither') prmT = c(0,0)
if (primeTypes[pt] == 'Both') prmT = c(1,1)
sim = simulate_nROUSE( presentations,
prmT, param );
par( mar=c( 3, 3, 2, .5 ) )
if ( pt == nrow(pltStr) ) par( mar=c(4,3,1,.5) )
yl = max( apply( sim$Activation, 2, max ) )
yl = c(0,ceiling(yl*10)/10)
xl = nrow(sim$Activation)
xl = c(0,xl)
plot( xl, yl, type = 'n', bty = 'n',
xlab = ' ', ylab = ' ',
xaxt = 'n', yaxt = 'n' )
abline(h=0,lwd=2)
mtext( primeTypes[pt], side = 2, cex = 1.25, line = -.75 )
if (pt==nrow(pltStr)) axis(1,round(seq(xl[1],xl[2],length=5)),
tick = F, line = -1, cex.axis = 1.4 )
if (pt == 1) {
xp = grconvertX( 1, from = "npc", to = "user" )*.1
if ( nrow(pltStr) == 4 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.25
if ( nrow(pltStr) == 3 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.2
if ( nrow(pltStr) == 2 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.1
legend( xp, yp, c('Prime','Target','Mask','Choices'),
fill = c('blue','orange','purple','grey'),
bty = 'n', cex= 1.4, horiz = T, xpd = T )
if ( nrow(pltStr) == 4 )
yp = grconvertY( 1, from = "npc", to = "user" )*.15
if ( nrow(pltStr) == 3 )
yp = grconvertY( 1, from = "npc", to = "user" )*.1
if ( nrow(pltStr) == 2 )
yp = grconvertY( 1, from = "npc", to = "user" )*.05
legend( xp, 0 - yp,
c('Target activation', 'Foil activation'),
fill = c('orange','red'),
bty = 'n', cex= 1.4, horiz = T, xpd = T )
}
for (i in 1:3) segments( cumsum(presentations[1:i])+.5, 0,
cumsum(presentations[1:i])+.5, yl[2],
col = 'grey80' )
lines( 1:nrow(sim$Activation),
sim$Activation[,1], lwd = 2, col = 'orange' )
lines( 1:nrow(sim$Activation),
sim$Activation[,2], lwd = 2, col = 'red' )
pck = numeric(2)
act = sim$Activation[ (pd+td+md+1):(pd+td+md+cd), ]
for (i in 1:2) {
pck[i] = which( act[,i] == max( act[,i] ) )
segments( pck[i] + (pd+td+md), 0,
pck[i] + (pd+td+md), act[pck[i],i], col = 'black',
lwd = 2 )
}
segments( 1, yl[2], pd, yl[2],
col = 'blue', lwd = 3 )
segments( pd+1, yl[2],
pd+td, yl[2],
col = 'orange', lwd = 3 )
segments( pd+td+1, yl[2],
pd+td+md, yl[2],
col = 'purple', lwd = 3 )
segments( pd+td+md+1, yl[2],
pd+td+md+cd, yl[2],
col = 'grey', lwd = 3 )
mu = diff(sim$Latencies[2:1])
sigma = sqrt( sum( exp( param['N']*sim$Latencies[1:2] ) ) )
val = seq( -4*sigma, 4*sigma, length = 1000 ) + mu
dn = dnorm( val, mu, sigma )
par( mar=c(1,.5,.5,3) )
plot( dn, val, type = 'n', xlab = ' ', ylab = ' ',
yaxt = 'n', xaxt = 'n', lwd = 2, bty = 'n' )
sel = val < 0
polygon( c(0, dn[sel], 0 ),
c( min(val[sel]), val[sel], max(val[sel]) ),
col = 'grey', border = 'NA' )
lines( dn, val, lwd = 2 )
abline( h = 0, lwd = 2, lty = 2 )
legend( 'bottom',
paste( round( sim$Latencies[3], 2 )*100, '%', sep = '' ),
bty = 'n', cex = 1.5 )
}
mtext('Activation',side=2,outer=T,cex=1.5,line=-1.75)
mtext('Duration (ms)',side=1,outer=T,cex=1.5,line=-1.5)
mtext('Predicted accuracy',side=4,outer=T,cex=1.5,line=-1.5)
}
rettopnivek/nROUSE documentation built on May 27, 2019, 5:55 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#----------------------------------------------#
# R code for plotting/demonstrations of nROUSE #
#----------------------------------------------#
# Index
# Lookup - 01: nROUSE_predictions
# Lookup - 02: nROUSE_demonstration
# Lookup - 01
#' Predictions from the nROUSE model
#'
#' Generates a plot of the predicted accuracy over different
#' prime durations and types for the nROUSE model (output can
#' be generated as well).
#'
#' @param primeDurations a vector of prime durations between
#' 17 and 2000 ms.
#' @param primeTypes a character vector with up to 4 types of prime:
#' 'Target', 'Foil', 'Neither', or 'Both'.
#' @param design a vector with the duration (in ms) of the target flash,
#' the target mask, and the choice options.
#' @param opt a logical vector indicating if the predicted accuracies
#' latencies should be returned, and if a plot should be generated.
#' @param prm an optional named vector for the parameters of the
#' nROUSE model, where...
#' \itemize{
#' \item Fe = the feedback scalar.
#' \item N = The noise multiplier.
#' \item L = The constant leak current.
#' \item D = The synaptic depletion rate.
#' \item R = The replenishment rate.
#' \item I = The inhibition constant.
#' \item Th = The noise multiplier.
#' \item Ta = The temporal attention parameter.
#' \item SV = The visual integration rate.
#' \item SO = The orthographic integration rate.
#' \item SS = The semantic integration rate.
#' }
#'
#'
#' @return A list consisting of a matrix of predicted accuracies by
#' prime duration and type, and an array with the predicted perceptual
#' identification latencies for the target and foil by prime duration
#' and prime type.
#'
#' @examples
#' # Default
#' nROUSE_predictions()
#'
#' # Specific conditions
#'
#' # Prime durations
#' pd = c(50,400)
#' # Prime types
#' pt = c('Target','Foil')
#' nROUSE_predictions( primeDurations = pd, primeTypes = pt )
#'
#' # Different timing for target stimulus flash and subsequent mask
#' nROUSE_predictions( design = c( 84, 416, 500 ) )
#'
#' # Suppress plotting and return predictions
#' pred = nROUSE_predictions( opt = c( T, F ) )
#'
#' # Different parameter values (Must be named)
#' prm = c( I = .5, Ta = .8, N = .025 )
#' nROUSE_predictions( prm = prm )
#'
#' @export
nROUSE_predictions = function( primeDurations = c( 17,50,150,400,2000 ),
primeTypes = c('Target','Foil','Neither','Both'),
design = c( 50, 450, 500 ),
opt = c( F, T ),
prm = NULL ) {
# Default parameters (Rieth & Huber, 2017)
param = c( Fe = .25, N = .0302, L = .15,
D = .324, R = .022, I = .9844,
Th = .15, Ta = 1.0, SV = .0294,
SO = .0609, SS = .015 )
# Check if parameter values should be changed
if ( length( prm ) > 0 ) {
selPar = names(prm)
for (i in 1:length(selPar)) {
chk = names(param) == selPar[i]
if ( sum(chk) == 1 )
param[ chk ] = prm[i]
}
}
# Sort prime duratiosn
primeDurations = sort( primeDurations );
# Check if prime types are admissable
chk = c('Target','Foil','Neither','Both')
if ( sum(primeTypes %in% chk) < length(primeTypes) )
stop('Check character vector for prime types.')
# Matrices to store output
Accuracy = matrix( NA, length( primeDurations ),
length( primeTypes ) )
colnames( Accuracy ) = primeTypes;
rownames( Accuracy ) = as.character( primeDurations )
Latency = array( NA, dim=c( length( primeDurations ),
length( primeTypes ),
2 ) )
# Simulate nROUSE
for ( pt in 1:length( primeTypes ) ) {
for ( pd in 1:length( primeDurations ) ) {
pres = c( primeDurations[pd], design )
if (primeTypes[pt] == 'Target') prmT = c(2,0)
if (primeTypes[pt] == 'Foil') prmT = c(0,2)
if (primeTypes[pt] == 'Neither') prmT = c(0,0)
if (primeTypes[pt] == 'Both') prmT = c(1,1)
sim = simulate_nROUSE( pres, prmT, param )
Accuracy[pd,pt] = sim$Latencies[3]
Latency[pd,pt,] = sim$Latencies[1:2]
}
}
# Plot results
if (opt[2]) {
clrs = c( Target = 'orange',
Foil = 'red',
Neither = 'blue',
Both = 'purple' )
xl = c( 10, max( primeDurations ) )
plot( log(xl), c(0,1), type = 'n', xaxt = 'n',
yaxt = 'n', xlab = ' ', ylab = ' ',
bty = 'l' )
xa = c(10,100,1000,5000)
axis( 1, log( xa[ xa <= max(xl) ] ),
xa[ xa <= max(xl) ], cex.axis = 1.5 )
mtext('Prime duration (log ms)',side=1,cex=1.5,
line = 2.25)
axis(2,seq(0,1,.25),cex.axis=1.5)
mtext('Accuracy',side=2,cex=1.5,
line = 2.25)
abline( h = .5, lty = 2, lwd = 2 )
mtext('nROUSE predictions', side = 3, line = 1,
cex = 1.5 )
whchClr = c()
for ( pt in 1:length( primeTypes ) ) {
clr = clrs[ names(clrs) %in% primeTypes[pt] ]
whchClr = c( whchClr, clr )
lines( log( primeDurations ), Accuracy[,pt], col = clr,
lwd = 2 )
points( log( primeDurations ), Accuracy[,pt], bg = clr,
pch = 21, cex = 1.2 )
}
legend('bottomleft',primeTypes, fill = whchClr, bty = 'n',
cex = 1.5 )
}
if (opt[1]) return( list( Accuracy = Accuracy, Latency = Latency ) )
}
# Lookup - 02
#' Demonstration of the nROUSE model.
#'
#' Generates a plot of the activation at the highest level of the
#' neural network (i.e. semantic-lexical) for the target and foil
#' representations under different prime types and an inputted
#' prime duration.
#'
#' @param primeTypes a character vector with up to 4 types of prime:
#' 'Target', 'Foil', 'Neither', or 'Both'.
#' @param design a vector with the duration (in ms) of the target flash,
#' the target mask, and the choice options.
#' @param prm an optional named vector for the parameters of the
#' nROUSE model, where...
#' \itemize{
#' \item Fe = the feedback scalar.
#' \item N = The noise multiplier.
#' \item L = The constant leak current.
#' \item D = The synaptic depletion rate.
#' \item R = The replenishment rate.
#' \item I = The inhibition constant.
#' \item Th = The noise multiplier.
#' \item Ta = The temporal attention parameter.
#' \item SV = The visual integration rate.
#' \item SO = The orthographic integration rate.
#' \item SS = The semantic integration rate.
#' }
#'
#' @details The user will receive a prompt to input a prime duration.
#' The function will then generate a plot showing the time course
#' of activation for the target/foil representations for the
#' set of prime types. The predicted accuracies and the difference
#' distribution from which these accuracies are derived are shown
#' on the right.
#'
#' @examples
#' \dontrun{
#' # Default
#' nROUSE_demonstration()
#'
#' # Fewer prime types
#' pt = c( 'Target', 'Foil' )
#' nROUSE_demonstration(primeTypes = pt)
#'
#' # Different timing for target stimulus flash and subsequent mask
#' nROUSE_demonstration(design=c(84,416,500))
#'
#' # Different parameter values (Must be named)
#' prm = c( I = .5, Ta = .8, N = .025 )
#' nROUSE_demonstration(prm=prm)
#' }
#'
#' @export
nROUSE_demonstration = function(primeTypes = c('Target','Foil',
'Neither','Both'),
design = c( 50, 450, 500 ),
prm = NULL ) {
# Default parameters
param = c( Fe = .25, N = .0302, L = .15,
D = .324, R = .022, I = .9844,
Th = .15, Ta = 1.0, SV = .0294,
SO = .0609, SS = .015 )
# Check if parameter values should be changed
if ( length( prm ) > 0 ) {
selPar = names(prm)
for (i in 1:length(selPar)) {
chk = names(param) == selPar[i]
if ( sum(chk) == 1 )
param[ chk ] = prm[i]
}
}
cat("Prime duration? (17 - 2000)","\n")
cat("Type number (in ms) and hit 'enter'","\n")
pd = scan(n = 1)
td = design[1]
md = design[2]
cd = design[3]
presentations = c( pd, design )
pltStr = c()
npt = length( primeTypes )
pltStr = matrix( rep( 1:length( primeTypes ), each = 8 ),
length( primeTypes ), 8, byrow = T )
pltStr = pltStr + 0:(length(primeTypes)-1)
pltStr = cbind( pltStr, pltStr[,1:2]+1 )
layout( pltStr )
for (pt in 1:length(primeTypes)) {
if (primeTypes[pt] == 'Target') prmT = c(2,0)
if (primeTypes[pt] == 'Foil') prmT = c(0,2)
if (primeTypes[pt] == 'Neither') prmT = c(0,0)
if (primeTypes[pt] == 'Both') prmT = c(1,1)
sim = simulate_nROUSE( presentations,
prmT, param );
par( mar=c( 3, 3, 2, .5 ) )
if ( pt == nrow(pltStr) ) par( mar=c(4,3,1,.5) )
yl = max( apply( sim$Activation, 2, max ) )
yl = c(0,ceiling(yl*10)/10)
xl = nrow(sim$Activation)
xl = c(0,xl)
plot( xl, yl, type = 'n', bty = 'n',
xlab = ' ', ylab = ' ',
xaxt = 'n', yaxt = 'n' )
abline(h=0,lwd=2)
mtext( primeTypes[pt], side = 2, cex = 1.25, line = -.75 )
if (pt==nrow(pltStr)) axis(1,round(seq(xl[1],xl[2],length=5)),
tick = F, line = -1, cex.axis = 1.4 )
if (pt == 1) {
xp = grconvertX( 1, from = "npc", to = "user" )*.1
if ( nrow(pltStr) == 4 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.25
if ( nrow(pltStr) == 3 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.2
if ( nrow(pltStr) == 2 )
yp = grconvertY( 1, from = "npc", to = "user" )*1.1
legend( xp, yp, c('Prime','Target','Mask','Choices'),
fill = c('blue','orange','purple','grey'),
bty = 'n', cex= 1.4, horiz = T, xpd = T )
if ( nrow(pltStr) == 4 )
yp = grconvertY( 1, from = "npc", to = "user" )*.15
if ( nrow(pltStr) == 3 )
yp = grconvertY( 1, from = "npc", to = "user" )*.1
if ( nrow(pltStr) == 2 )
yp = grconvertY( 1, from = "npc", to = "user" )*.05
legend( xp, 0 - yp,
c('Target activation', 'Foil activation'),
fill = c('orange','red'),
bty = 'n', cex= 1.4, horiz = T, xpd = T )
}
for (i in 1:3) segments( cumsum(presentations[1:i])+.5, 0,
cumsum(presentations[1:i])+.5, yl[2],
col = 'grey80' )
lines( 1:nrow(sim$Activation),
sim$Activation[,1], lwd = 2, col = 'orange' )
lines( 1:nrow(sim$Activation),
sim$Activation[,2], lwd = 2, col = 'red' )
pck = numeric(2)
act = sim$Activation[ (pd+td+md+1):(pd+td+md+cd), ]
for (i in 1:2) {
pck[i] = which( act[,i] == max( act[,i] ) )
segments( pck[i] + (pd+td+md), 0,
pck[i] + (pd+td+md), act[pck[i],i], col = 'black',
lwd = 2 )
}
segments( 1, yl[2], pd, yl[2],
col = 'blue', lwd = 3 )
segments( pd+1, yl[2],
pd+td, yl[2],
col = 'orange', lwd = 3 )
segments( pd+td+1, yl[2],
pd+td+md, yl[2],
col = 'purple', lwd = 3 )
segments( pd+td+md+1, yl[2],
pd+td+md+cd, yl[2],
col = 'grey', lwd = 3 )
mu = diff(sim$Latencies[2:1])
sigma = sqrt( sum( exp( param['N']*sim$Latencies[1:2] ) ) )
val = seq( -4*sigma, 4*sigma, length = 1000 ) + mu
dn = dnorm( val, mu, sigma )
par( mar=c(1,.5,.5,3) )
plot( dn, val, type = 'n', xlab = ' ', ylab = ' ',
yaxt = 'n', xaxt = 'n', lwd = 2, bty = 'n' )
sel = val < 0
polygon( c(0, dn[sel], 0 ),
c( min(val[sel]), val[sel], max(val[sel]) ),
col = 'grey', border = 'NA' )
lines( dn, val, lwd = 2 )
abline( h = 0, lwd = 2, lty = 2 )
legend( 'bottom',
paste( round( sim$Latencies[3], 2 )*100, '%', sep = '' ),
bty = 'n', cex = 1.5 )
}
mtext('Activation',side=2,outer=T,cex=1.5,line=-1.75)
mtext('Duration (ms)',side=1,outer=T,cex=1.5,line=-1.5)
mtext('Predicted accuracy',side=4,outer=T,cex=1.5,line=-1.5)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.