keep <- keep & tmp.data$Rater != "MCC" & tmp.data$rated != "MCC"
#initial <- as.formula(paste("rating~ 0+",paste(colnames[c(1:5,7:11)],collapse="+")))
vars4 <- vars3[keep,]
initial <- as.formula(paste("rating~ 0+",paste(varnames,collapse="+")))
start <- lm(formula=initial,data=vars4)
full <- lm(formula=full.model,data=vars4)
base <- lm(formula=base.model,data=vars4)
x3 <- stepAIC(start,scope=c(lower=lower.model,upper=upper.model),trace=F,direction="both")
cont <- 1
m1 <- x3
lastlogLike <- sum(log(dnorm(m1$residuals,0,summary(m1)$sigma)))
cluster <- rep(1,length(m1$residuals))
## A base rate distribution over the ratings
#m2 <-  table(vars4$rating)/nrow(vars4)
m2 <- rep(1,11)/11
#m2 <- round(m2+1)/11##make it uniform actually.
step <- 1
while(cont)
{
modelLike <- data.frame(m1Like = dnorm(m1$residuals,0,summary(m1)$sigma),
m2Like = as.vector(m2)[vars4$rating+1])
cluster <-  apply(modelLike,1, which.max)
modelLike$maxlike <- modelLike[cbind(1:nrow(modelLike),cluster)]
logLike <- sum(log(modelLike$maxlike))
if(logLike > lastlogLike)
{
##row, re-run the step model using m1 as the starting point, but
## with mstep==1 as the filter.
start <- lm(formula=formula(m1),data=vars4,weights= ((cluster==1)+0))
m1 <- stepAIC(start,scope=c(lower=lower.model,upper=upper.model),trace=F,direction="both")#,k=log(nrow(r.ls.long)))
lastlogLike <- logLike
cat("Step: [",step,"]  log likelihood:",lastlogLike, "->",logLike,"\n")
} else{
## we have not improved, so end the model cycles
cont <- 0
}
step <- step + 1
}
fit <- coef(m1)
fit[is.na(fit)] <- 0
rated <- fit[1:5]
rater <- fit[6:10]
daycoef <- fit[23:147]
daynames <- names(daycoef)
byday <- data.frame(Day=sapply(daynames, function(x){as.numeric(substr(strsplit(x,'[.]')[[1]][[2]],4,6))}),
Rated = sapply(daynames, function(x){strsplit(x,'[.]')[[1]][[1]]}),
value = daycoef)
byday$ratedval <- rep(rated,each=25) ##hand-code 25 times here.
byday$adjusted <- byday$ratedval + byday$value
meanrated <- mean(rated)
meanrater <- mean(rater)
meanday <- mean(byday$value)
grandmean <- mean(vars3$rating[keep])
int2 <- grandmean
tmp <- data.frame(Member=substr(names(rated),7,9),rating=rated)
p.adapt.rated <-ggplot(tmp,aes(Member,y=rating)) + geom_bar(stat="identity")+theme_bw() + ggtitle(paste("Mean rating of each member by others:",dimension))
barplot(rated,names=roles[1:5],main=paste("Mean rating of each member by others:",dimension),col="gold")
abline(int2,0)
barplot(rater, names=roles[1:5],main=paste("Bias of each rater:",dimension), col="cadetblue4")
abline(int2,0)
ggplot(byday,aes(x=Day,y=value,group=Rated,color=Rated)) + geom_point()+geom_line() + ggtitle(label=paste(dimension,"per team member over time")) + theme_bw()
ggplot(byday,aes(x=Day,y=adjusted,group=Rated,color=Rated)) + geom_point()+geom_line() + ggtitle(label=paste(dimension,"per team member over time (re-meaned)")) + theme_bw()
extra <- fit[c(11:21)]
pairs.adapt <- data.frame(pair=names(extra), adjustment= extra)
ggplot(pairs.adapt,aes(pair,adjustment)) + geom_bar(stat="identity") + theme_bw() + ggtitle(label=paste("Pairwise deviations from model:",dimension))
allout3 <- r.adapt.long[keep,]
allout3$value[allout3$Rater==allout3$rated] <- NA
allout3$fittedbase <- base$fitted.values
allout3$fitted <- m1$fitted.values
allout3$fittedfull <- full$fitted.values
allout3$cluster <- as.factor(cluster)
ggplot(allout3,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (data)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fittedbase,group=cond,col=Rater)) + geom_line() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Baseline model)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fittedfull,group=cond,col=Rater)) + geom_line() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Full model)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fitted,group=cond,col=Rater)) + geom_line() +
theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Simplified model)",sep=""))  +  geom_point()+
geom_point(data=allout3[allout3$cluster==2,],aes(x=Day,y=value),color="red",shape=11)  +
facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=value,y=fitted,col=Rater)) + geom_point(aes(shape=cluster)) + theme_minimal() + ggtitle(label=paste("Data vs. fitted model:",dimension))
table(cluster)
tmp <- r.adapt.long
tmp <- tmp[tmp$Rater != tmp$rated,]
tmp <- tmp[tmp$rated != "MCC",]
ggplot(tmp,aes(x=Day,y=value,group=cond,col=rated)) + geom_line() + theme_bw()  + ylim(0,10) + xlim(0,240)+ ggtitle(label=paste("Outlier detection | ",dimension,": Self-ratings vs. how member rated others",sep="") )+
geom_point(data=allout3[cluster==2,],aes(x=Day,y=value),size=1.5,colour="red",shape=11)    +
facet_wrap(~Rater,ncol=5)
#  geom_line(data=self.ls,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
aggregate(allout3$rated,list(Dimension=allout3$Dimension,
rater=allout3$Rater,keep=allout3$cluster),length)
dimension <- "Shared Mental Model"
tmp.data <- r.smm.long
vars3 <- vars
vars3$rating <- tmp.data$value
vars3$CDR.day <- as.character(vars3$Day)
vars3$CS.day <-  as.character(vars3$Day)
vars3$FE.day <-  as.character(vars3$Day)
vars3$MS1.day <- as.character(vars3$Day)
vars3$MS3.day <- as.character(vars3$Day)
vars3$CDR.day[vars3$rated.CDR==0] <- 0
vars3$CS.day[vars3$rated.CS==0] <- 0
vars3$FE.day[vars3$rated.FE==0] <- 0
vars3$MS1.day[vars3$rated.MS1==0] <- 0
vars3$MS3.day[vars3$rated.MS3==0] <- 0
varnames <- c(colnames,"CDR.day","CS.day","FE.day","MS1.day","MS3.day")
upper.model <- as.formula(paste("~ 0+",paste(varnames,collapse="+")))
lower.model <- as.formula(paste("~ 0+",paste(varnames[c(1:5,7:11, (length(varnames) - (4:0))) ],collapse="+")))
full.model <-  as.formula(paste("rating ~ 0+",paste(varnames,collapse="+")))
base.model <-  as.formula(paste("rating ~ 0+",paste(varnames[c(1:5,7:11, (length(varnames) - (4:0))) ],collapse="+")))
keep <- !is.na(vars3$rating)
keep <- keep & !(tmp.data$Rater == tmp.data$rated)
keep <- keep & tmp.data$Rater != "MCC" & tmp.data$rated != "MCC"
#initial <- as.formula(paste("rating~ 0+",paste(colnames[c(1:5,7:11)],collapse="+")))
vars4 <- vars3[keep,]
initial <- as.formula(paste("rating~ 0+",paste(varnames,collapse="+")))
start <- lm(formula=initial,data=vars4)
full <- lm(formula=full.model,data=vars4)
base <- lm(formula=base.model,data=vars4)
x3 <- stepAIC(start,scope=c(lower=lower.model,upper=upper.model),trace=F,direction="both")
cont <- 1
m1 <- x3
lastlogLike <- sum(log(dnorm(m1$residuals,0,summary(m1)$sigma)))
cluster <- rep(1,length(m1$residuals))
## A base rate distribution over the ratings
#m2 <-  table(vars4$rating)/nrow(vars4)
m2 <- rep(1,11)/11
#m2 <- round(m2+1)/11##make it uniform actually.
step <- 1
while(cont)
{
modelLike <- data.frame(m1Like = dnorm(m1$residuals,0,summary(m1)$sigma),
m2Like = as.vector(m2)[vars4$rating+1])
cluster <-  apply(modelLike,1, which.max)
modelLike$maxlike <- modelLike[cbind(1:nrow(modelLike),cluster)]
logLike <- sum(log(modelLike$maxlike))
if(logLike > lastlogLike)
{
##row, re-run the step model using m1 as the starting point, but
## with mstep==1 as the filter.
start <- lm(formula=formula(m1),data=vars4,weights= ((cluster==1)+0))
m1 <- stepAIC(start,scope=c(lower=lower.model,upper=upper.model),trace=F,direction="both")#,k=log(nrow(r.ls.long)))
lastlogLike <- logLike
cat("Step: [",step,"]  log likelihood:",lastlogLike, "->",logLike,"\n")
} else{
## we have not improved, so end the model cycles
cont <- 0
}
step <- step + 1
}
cluster.smm <-  apply(modelLike,1, which.max)
fit <- coef(m1)
fit[is.na(fit)] <- 0
rated <- fit[1:5]
rater <- fit[6:10]
daycoef <- fit[22:146]
daynames <- names(daycoef)
byday <- data.frame(Day=sapply(daynames, function(x){as.numeric(substr(strsplit(x,'[.]')[[1]][[2]],4,6))}),
Rated = sapply(daynames, function(x){strsplit(x,'[.]')[[1]][[1]]}),
value = daycoef)
byday$ratedval <- rep(rated,each=25) ##hand-code 25 times here.
byday$adjusted <- byday$ratedval + byday$value
meanrated <- mean(rated)
meanrater <- mean(rater)
meanday <- mean(byday$value)
grandmean <- mean(vars3$rating[keep])
int2 <- grandmean
tmp <- data.frame(Member=substr(names(rated),7,9),rating=rated)
p.smm.rated <-ggplot(tmp,aes(Member,y=rating)) + geom_bar(stat="identity")+theme_bw() + ggtitle(paste("Mean rating of each member by others:",dimension))
barplot(rated,names=roles[1:5],main=paste("Mean rating of each member by others:",dimension),col="gold")
abline(int2,0)
barplot(rater, names=roles[1:5],main=paste("Bias of each rater:",dimension), col="cadetblue4")
abline(int2,0)
ggplot(byday,aes(x=Day,y=value,group=Rated,color=Rated)) + geom_point()+geom_line() + ggtitle(label=paste(dimension,"per team member over time")) + theme_bw()
ggplot(byday,aes(x=Day,y=adjusted,group=Rated,color=Rated)) + geom_point()+geom_line() + ggtitle(label=paste(dimension,"per team member over time (re-meaned)")) + theme_bw()
extra <- fit[c(11:20)]
pairs.smm <- data.frame(pair=names(extra), adjustment= extra)
ggplot(pairs.smm,aes(pair,adjustment)) + geom_bar(stat="identity") + theme_bw() + ggtitle(label=paste("Pairwise deviations from model:",dimension))
allout3 <- r.smm.long[keep,]
allout3$value[allout3$Rater==allout3$rated] <- NA
allout3$fittedbase <- base$fitted.values
allout3$fitted <- m1$fitted.values
allout3$fittedfull <- full$fitted.values
allout3$cluster <- as.factor(cluster)
ggplot(allout3,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (data)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fittedbase,group=cond,col=Rater)) + geom_line() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Baseline model)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fittedfull,group=cond,col=Rater)) + geom_line() + theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Full model)",sep=""))  +   facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=Day,y=fitted,group=cond,col=Rater)) + geom_line() +
theme_bw()  + ylim(0,11) + xlim(0,240)+ ggtitle(label=paste(dimension,": As rated by others (Simplified model)",sep=""))  +  geom_point()+
geom_point(data=allout3[allout3$cluster==2,],aes(x=Day,y=value),color="red",shape=11)  +
facet_wrap(~rated,ncol=6)
ggplot(allout3,aes(x=value,y=fitted,col=Rater)) + geom_point(aes(shape=cluster)) + theme_minimal() + ggtitle(label=paste("Data vs. fitted model:",dimension))
table(cluster)
tmp <- r.smm.long
tmp <- tmp[tmp$Rater != tmp$rated,]
tmp <- tmp[tmp$rated != "MCC",]
ggplot(tmp,aes(x=Day,y=value,group=cond,col=rated)) + geom_line() + theme_bw()  + ylim(0,10) + xlim(0,240)+ ggtitle(label=paste("Outlier detection | ",dimension,": Self-ratings vs. how member rated others",sep="") )+
geom_point(data=allout3[cluster==2,],aes(x=Day,y=value),size=1.5,colour="red",shape=11)    +
facet_wrap(~Rater,ncol=5)
#  geom_line(data=self.ls,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
aggregate(allout3$rated,list(Dimension=allout3$Dimension,
rater=allout3$Rater,keep=allout3$cluster),length)
library(gridExtra)
grid.arrange(p.trust.rated,
p.torient.rated,
p.comm.rated,
p.backup.rated,
p.monitor.rated,
p.adapt.rated,
p.ls.rated,
p.smm.rated,
nrow=4)
p.trust.rated
p.torient.rated
p.comm.rated
p.backup.rated
p.monitor.rated
p.adapt.rated
p.ls.rated
library(ggnetwork)
library(network)
library(sna)
pairs.smm$dimension <- "SMM"
pairs.comm$dimension <- "COMM"
pairs.ls$dimension <- "Leadership"
pairs.trust$dimension <- "Trust"
pairs.backup$dimension <- "Backup"
pairs.adapt$dimension <- "Adaptability"
pairs.torient$dimension <- "Team orientation"
pairs.monitor$dimension <- "Performance monitoring"
pairs.all <- rbind(pairs.smm,pairs.comm,pairs.ls,pairs.trust,
pairs.backup,pairs.adapt,pairs.torient,pairs.monitor)
members <-sapply(pairs.all$pair, function(x){strsplit(x,'[.]')[[1]]})
pairs.all$rated <- factor( members[1,],levels=c("CDR","CS","FE","MS1","MS3"))
pairs.all$rater <- factor(members[2,],levels=c("CDR","CS","FE","MS1","MS3"))
pairs.tab <- tapply(pairs.all$adjustment,list(pairs.all$rater,pairs.all$rated,pairs.all$dimension),
function(x){tmp <- mean(x)})
pairs.tab[is.na(pairs.tab)] <- 0
library(tidyr)
##convert back to a data frame
pairs.all2 <-as.data.frame.table(pairs.tab)
colnames(pairs.all2) <- c("rater","rated","dimension","adjustment")
pairs.aug <- rbind(pairs.all,
data.frame(pair=rep("",40),
adjustment=rep(NA,40),
dimension=rep(unique(pairs.all$dimension),5),
rated=rep(levels(pairs.all$rated),each=8),
rater=rep(levels(pairs.all$rated),each=8)))
ggplot(pairs.all2,
aes(x=rated,y=rater,fill=adjustment))+
geom_tile(lwd=.5,color="grey80",linetype=1)+theme_minimal() +
coord_fixed()+facet_wrap(~dimension,ncol=4) + scale_fill_gradient2(low="red",high=("white"),midpoint=-1)
for(i in unique(pairs.all$dimension))
{
tmp0  <- pairs.all[pairs.all$dimension==i,]
tmp <- as.network(data.frame(from=tmp0$rater,
to= tmp0$rated,
weight=abs(tmp0$adjustment)),
vertices = data.frame(member=c("CDR","CS","FE","MS1","MS3"),xpos=1:5,ypos=c(1,1,2,2,1)))
p <- ggplot(ggnetwork(tmp),aes(x=x,y=y,xend=xend,yend=yend), layout="circle") +
geom_edges(size=.8,aes(color=weight) ,arrow = arrow(length = unit(.5, "cm"), angle=20,type = "closed"))+
geom_nodes(size=13,color="grey")+geom_nodetext(aes(label=vertex.names))+
theme_blank() + ggtitle(i)
print(p)
}
ratings$month <- as.factor(floor(ratings$Week /8.999)+1)
r.trust <-dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Trust"))
r.torient <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Torient"))
r.comm <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Comm"))
r.ls <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("LS"))
r.backup <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Backup"))
r.monitor <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Monitor"))
r.adapt <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("Adapt"))
r.smm <- dplyr::select(ratings,Day,Week,month,TaskType,Rater, starts_with("SMM"))
r.trust.long <- melt(r.trust,measure.vars=c("Trust.CDR","Trust.FE","Trust.MS1","Trust.MS3","Trust.CS","Trust.MCC"))
r.trust.long$cond <- paste(r.trust.long$variable,r.trust.long$Rater)
r.trust.long$rated <- substr(r.trust.long$variable,7,12)
r.trust.long$rated <- factor(r.trust.long$rated,levels=roles)
r.trust.long$Dimension <- "Mutual Trust"
self.trust <- subset(r.trust.long,Rater==rated)
self.trust$Rater <- "SELF"
r.torient.long <- melt(r.torient,measure.vars=c("Torient.CDR","Torient.FE","Torient.MS1","Torient.MS3","Torient.CS","Torient.MCC"))
r.torient.long$cond <- paste(r.torient.long$variable,r.torient.long$Rater)
r.torient.long$rated <- substr(r.torient.long$variable,9,14)
r.torient.long$rated <- factor(r.torient.long$rated,levels=roles)
r.torient.long$Dimension <- "Team orientations"
self.torient <- subset(r.torient.long,Rater==rated)
self.torient$Rater <- "SELF"
r.comm.long <- melt(r.comm,measure.vars=c("Comm.CDR","Comm.FE","Comm.MS1","Comm.MS3","Comm.CS","Comm.MCC"))
r.comm.long$cond <- paste(r.comm.long$variable,r.comm.long$Rater)
r.comm.long$rated <- substr(r.comm.long$variable,6,12)
r.comm.long$rated <- factor(r.comm.long$rated,levels=roles)
r.comm.long$Dimension <- "Communications"
self.comm <- subset(r.comm.long,Rater==rated)
self.comm$Rater <- "SELF"
r.ls.long <- melt(r.ls,measure.vars=c("LS.CDR","LS.FE","LS.MS1","LS.MS3","LS.CS","LS.MCC"))
r.ls.long$cond <- paste(r.ls.long$variable,r.ls.long$Rater)
r.ls.long$rated <- substr(r.ls.long$variable,4,12)
r.ls.long$rated <- factor(r.ls.long$rated,levels=roles)
r.ls.long$Dimension <- "Leadership"
self.ls <- subset(r.ls.long,Rater==rated)
self.ls$Rater <- "SELF"
ls.fe <- subset(r.ls.long,rated=="FE")
r.backup.long <- melt(r.backup,measure.vars=c("Backup.CDR","Backup.FE","Backup.MS1","Backup.MS3","Backup.CS","Backup.MCC"))
r.backup.long$cond <- paste(r.backup.long$variable,r.backup.long$Rater)
r.backup.long$rated <- substr(r.backup.long$variable,8,14)
r.backup.long$rated <- factor(r.backup.long$rated,levels=roles)
r.backup.long$Dimension <- "Backup behavior"
self.backup <- subset(r.backup.long,Rater==rated)
self.backup$Rater <- "SELF"
r.monitor.long <- melt(r.monitor,measure.vars=c("Monitor.CDR","Monitor.FE","Monitor.MS1","Monitor.MS3","Monitor.CS","Monitor.MCC"))
r.monitor.long$cond <- paste(r.monitor.long$variable,r.monitor.long$Rater)
r.monitor.long$rated <- substr(r.monitor.long$variable,9,15)
r.monitor.long$rated <- factor(r.monitor.long$rated,levels=roles)
r.monitor.long$Dimension <- "Performance monitoring"
self.monitor <- subset(r.monitor.long,Rater==rated)
self.monitor$Rater <- "SELF"
r.adapt.long <- melt(r.adapt,measure.vars=c("Adapt.CDR","Adapt.FE","Adapt.MS1","Adapt.MS3","Adapt.CS","Adapt.MCC"))
r.adapt.long$cond <- paste(r.adapt.long$variable,r.adapt.long$Rater)
r.adapt.long$rated <- substr(r.adapt.long$variable,7,15)
r.adapt.long$rated <- factor(r.adapt.long$rated,levels=roles)
r.adapt.long$Dimension <- "Adaptability"
self.adapt <- subset(r.adapt.long,Rater==rated)
self.adapt$Rater <- "SELF"
r.smm.long <- melt(r.smm,measure.vars=c("SMM.CDR","SMM.FE","SMM.MS1","SMM.MS3","SMM.CS","SMM.MCC"))
r.smm.long$cond <- paste(r.smm.long$variable,r.smm.long$Rater)
r.smm.long$rated <- substr(r.smm.long$variable,5,15)
r.smm.long$rated <- factor(r.smm.long$rated,levels=roles)
r.smm.long$Dimension <- "Shared mental model"
self.smm <- subset(r.smm.long,Rater==rated)
self.smm$Rater <- "SELF"
ggplot(r.trust.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw() + ylim(0,10) + xlim(0,240)+
ggtitle(label="Trust: As rated by others vs. self")+ geom_jitter(width=.5)   + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.trust,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.torient.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10)+ xlim(0,240)+
ggtitle(label="Team Orientation: As rated by others vs. self")+ geom_jitter(width=.5)  + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.torient,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.comm.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10)+ xlim(0,240)+ ggtitle(label="Communications: As rated by others vs. self")+ geom_jitter(width=.5) + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.comm,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.ls.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10) + xlim(0,240)+ ggtitle(label="Leadership: As rated by others vs. self")+ geom_jitter(width=.5)  + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.ls,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.backup.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10) + xlim(0,240)+
ggtitle(label="Back-up behavior: As rated by others vs. self")+ geom_jitter(width=.5)  + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.backup,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.monitor.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10) + xlim(0,240)+ ggtitle(label="Performance monitoring: As rated by others vs. self")+ geom_jitter(width=.5)  + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.monitor,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.adapt.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10) + xlim(0,240)+ ggtitle(label="Adaptability: As rated by others vs. self")+ geom_jitter(width=.5)   + facet_wrap(~rated,ncol=6)  +
geom_line(data=self.adapt,aes(x=Day,y=value),col="grey20",alpha=.4,size=2)
ggplot(r.smm.long,aes(x=Day,y=value,group=cond,col=Rater)) + geom_line() +geom_point() + theme_bw()  + ylim(0,10) + xlim(0,240)+
ggtitle(label="Shared mental model: As rated by others vs. self") + geom_jitter(width=.5)  + facet_wrap(~rated,ncol=6) +
geom_line(data=self.smm,aes(x=Day,y=value),col="grey20",alpha=.3,size=2)
#title self-image versus others' ratings
# self-rating compared to how others rate person
#vs
aggregate(allout3$rated,list(Dimension=allout3$Dimension,
rater=allout3$Rater,rated=allout3$rated,keep=allout3$cluster),length)
setwd("~/Dropbox/courses/5210-2022/web-5210/psy5210/Projects/Chapter4")
dat2 <- read.csv("pooled-pursuitrotor.csv")
dim(dat2)
head(dat2)
table(dat2$subnum)
table(dat2$subnum,dat2$trial)
##select on a specific trial/participant
tmp <- dat2[dat2$trial==1&dat2$subnum==12818,]
dim(dat2)
dim(tmp)
table(tmp$tdiff)
par(mfrow=c(1,3))
plot(tmp$timeElapsed, main= paste("Mean:" , round(mean(tmp$tdiff),3)))
plot(tmp$tdiff)
hist(tmp$tdiff)
abline(v=mean(tmp$tdiff))
matplot(cbind(tmp$targX,tmp$targY),
type="l",bty="L",las=1,lty=2,
xlab="Time step",ylab="Pixel location",
ylim=c(0,1500))
legend(550,1500,c("X position","Y position"),lty=1:2,col=1:2)
matplot(cbind(tmp$mouseX,tmp$mouseY),add=T,
type="l",las=1,lty=1)
xmid <- mean(tmp$targX) ## find the x center
ymid <- mean(tmp$targY) ## find they center
plot(tmp$targX-xmid,ymid-tmp$targY,pch=1,cex=.2,
xlim=c(-300,300),ylim=c(-300,300),
xlab="Horizontal pixel",ylab="Vertical pixel")
points(tmp$mouseX-xmid,ymid-tmp$mouseY,type="l",col="grey")
segments(tmp$targX-xmid,ymid-tmp$targY,tmp$mouseX-xmid,
ymid-tmp$mouseY,col="grey")
for(i in 1:nrow(tmp))
{
if(tmp[i,]$ontarget)
{
points(tmp$mouseX[i]-xmid,ymid-tmp$mouseY[i],col="red",cex=.3)
}else{
points(tmp$mouseX[i]-xmid,ymid-tmp$mouseY[i],col="grey",cex=.3)
}
}
plottrial <- function(tmp,header="")
{
xmid <- mean(tmp$targX)
ymid <- mean(tmp$targY)
meanoffset <- round(mean(tmp$diff),2)
plot(tmp$targX-xmid,ymid-tmp$targY,pch=1,cex=.2,
xlim=c(-300,300),ylim=c(-300,300),
#       main=paste(header,"\n","mean offset =", meanoffset),
xlab="",ylab="",axes=F)
#mousex, mousey specify the mouse coordinates: Plot them to:
points(tmp$mouseX-xmid,ymid-tmp$mouseY,type="l",col="grey")
segments(tmp$targX-xmid,ymid-tmp$targY,tmp$mouseX-xmid,ymid-tmp$mouseY,col="grey")
##I'll use a faster points method than we did above:
points(tmp$mouseX-xmid,ymid-tmp$mouseY,
col=c("grey","red")[tmp$ontarget+1],cex=.3)
}
plottrial(tmp,"test")
sub1 <- dat2[dat2$subnum==12887,]
for(i in levels(sub1$trial))
{
print(i)
}
levels(sub1$trial)
names(table(sub1$trial))
levels(as.factor(sub1$trial))
for(i in levels(as.factor(sub1$trial)))
{
print(i)
}
sub1 <- dat2[dat2$subnum==12887,]
par(mfrow=c(2,2))
for(i in levels(as.factor(sub1$trial)))
{
tmp <- sub1[sub1$trial==i,]
plottrial(tmp)
}
tapply(dat2$ontarget,list(dat2$sub,dat2$trial),mean)
for(sub in levels(as.factor(dat2$subnum)))
{
par(mfrow=c(2,2))
tmp <- dat2[dat2$subnum==sub,]
for(trial in levels(as.factor(tmp$trial)))
{
print(paste("plotting subject:", sub, " trial: ", trial))
tmp2 <- tmp[tmp$trial==trial,]
plottrial(tmp2,header=paste("plotting subject:", sub, " trial: ", trial))
}
}
pr.agg <- aggregate(dat2$diff,list(dat2$subnum,dat2$trial),mean)
par(mfrow=c(2,2))
hist(pr.agg$x,xlab="Mean deviation")
hist(pr.agg$x,xlab= "Mean deviation", col="navy",breaks=20)
hist(pr.agg$x,xlab= "Mean deviation", col="red",breaks=c(0:40))
##plots density instead of counts:
hist(pr.agg$x,xlab= "Mean deviation", col="darkgreen",breaks=c(0,10:20,25,30,40),freq=F )
View(InsectSprays)
## Notice this is in long data format.
## Not included in book
boxplot(InsectSprays$count~InsectSprays$spray)
##figures here will be 3x4 inches
set.seed(100)
data2 <- data.frame(q1=sample(letters[1:10],100,replace=T),
q2=sample(letters[1:10],100,replace=T),
q3=sample(letters[1:10],100,replace=T),
q4=sample(letters[1:10],100,replace=T),
q5=sample(letters[1:10],100,replace=T))
datatable2<-apply(data2,2,table)
lf <- c(8.167,1.492,2.782,4.253,12.702,2.228,2.015,6.094,
6.966,0.153,0.772,4.025,2.406,6.749,7.507,1.929,
0.095,5.987,6.327,9.056,2.758,0.978,2.36,0.15,1.974,0.074)/100
pts <- c(1,3,3,2,1,4,2,4,1,8,5,1,3,1,1,3,10,1,1,1,1,4,4,8,4,10)
tiles <- c(9,2,2,4,12,2,3,2,9,1,1,4,2,6,8,2,1,6,4,6,4,2,2,1,2,1)
lf.table <- data.frame(LETTERS,
freq=lf,
points=pts,
ntiles=tiles)
scoreme <- function(word)
{
lets <- strsplit(splus2R::upperCase(word),"")[[1]]
data <- matrix(0,ncol=4,nrow=length(lets))
for(i in 1:length(lets))
{
index <- which(lets[i]==LETTERS)
data[i,1] <- lf.table$freq[index]
data[i,2] <- lf.table$points[index]
data[i,3] <- lf.table$ntiles[index]
}
list(suminvfreq= sum(1/data[,1]),
points=sum(data[,2]),
meantiles=mean(data[,3]),
length=length(lets))
}
horses <- scoreme("HORSES")
horses
print(horses$points)
test <- read.table(text='rank word frequency
1081  CUP      1441306
2310  FOUND     573305
5285  BUTTERFLY 171410
7371 	brew       94904
11821 CUMBERSOME 39698
17331 useable	 	 17790
18526 WHITTLE    15315
25416 SPINY       7207
27381 uppercase 	 	5959
37281 	  halfnaked 	 	2459
47381 	  bellhop 	 	1106
57351 	  tetherball 	 	425
7309 	    attic 	 	2711
17311 	  tearful 	 	542
27303 	  tailgate 	198
37310 	  hydraulically 		78
47309 	  unsparing 	 	35
57309 	  embryogenesis 	22 ',header=T)[,c(2,1,3)] ##reorder colums
test$meantiles <- NA
test$suminvfreq <- NA
test$points <- NA
test$length <- NA