I tried to use princomp() and principal() to do PCA in R with data set USArressts. However, I got two different results for loadings/rotaion and scores.
First, I centered and normalised the original data frame so it is easier to compare the outputs.
library(psych)
trans_func <- function(x){
x <- (x-mean(x))/sd(x)
return(x)
}
A <- USArrests
USArrests <- apply(USArrests, 2, trans_func)
princompPCA <- princomp(USArrests, cor = TRUE)
principalPCA <- principal(USArrests, nfactors=4 , scores=TRUE, rotate = "none",scale=TRUE)
Then I got the results for the loadings and scores using the following commands:
princompPCA$loadings
principalPCA$loadings
Could you please help me to explain why there is a difference? and how can we interprete these results?
At the very end of the help document of ?principal:
"The eigen vectors are rescaled by the sqrt of the eigen values to produce the component loadings more typical in factor analysis."
So principal returns the scaled loadings. In fact, principal produces a factor model estimated by the principal component method.
In 4 years, I would like to provide a more accurate answer to this question. I use iris data as an example.
data = iris[, 1:4]
First, do PCA by the eigen-decomposition
eigen_res = eigen(cov(data))
l = eigen_res$values
q = eigen_res$vectors
Then the eigenvector corresponding to the largest eigenvalue is the factor loadings
q[,1]
We can treat this as a reference or the correct answer. Now we check the results by different r functions.
First, by function 'princomp'
res1 = princomp(data)
res1$loadings[,1]
# compare with
q[,1]
No problem, this function actually just return the same results as 'eigen'. Now move to 'principal'
library(psych)
res2 = principal(data, nfactors=4, rotate="none")
# the loadings of the first PC is
res2$loadings[,1]
# compare it with the results by eigendecomposition
sqrt(l[1])*q[,1] # re-scale the eigen vector by sqrt of eigen value
You may find they are still different. The problem is the 'principal' function does eigendecomposition on the correlation matrix by default. Note: PCA is not invariant with rescaling the variables. If you modify the code as
res2 = principal(data, nfactors=4, rotate="none", cor="cov")
# the loadings of the first PC is
res2$loadings[,1]
# compare it with the results by eigendecomposition
sqrt(l[1])*q[,1] # re-scale the eigen vector by sqrt of eigen value
Now, you will get the same results as 'eigen' and 'princomp'.
Summarize:
If you want to do PCA, you'd better apply 'princomp' function.
PCA is a special case of the Factor model or a simplified version of the factor model. It is just equivalent to eigendecomposition.
We can apply PCA to get an approximation of a factor model. It doesn't care about the specific factors, i.e. epsilons in a factor model. So, if you change the number of factors in your model, you will get the same estimations of the loadings. It is different from the maximum likelihood estimation.
If you are estimating a factor model, you'd better use 'principal' function, since it provides more functions, like rotation, calculating the scores by different methods, and so on.
Rescale the loadings of a PCA model doesn't affect the results too much. Since you still project the data onto the same optimal direction, i.e. maximize the variation in the resulting PC.
ev <- eigen(R) # R is a correlation matrix of DATA
ev$vectors %*% diag(ev$values) %*% t(ev$vectors)
pc <- princomp(scale(DATA, center = F, scale = T),cor=TRUE)
p <-principal(DATA, rotate="none")
#eigen values
ev$values^0.5
pc$sdev
p$values^0.5
#eigen vectors - loadings
ev$vectors
pc$loadings
p$weights %*% diag(p$values^0.5)
pc$loading %*% diag(pc$sdev)
p$loadings
#weights
ee <- diag(0,2)
for (j in 1:2) {
for (i in 1:2) {
ee[i,j] <- ev$vectors[i,j]/p$values[j]^0.5
}
};ee
#scores
s <- as.matrix(scale(DATA, center = T, scale = T)) %*% ev$vectors
scale(s)
p$scores
scale(pc$scores)
I am working with the cumulative emergence of flies over time (taken at irregular intervals) over many summers (though first I am just trying to make one year work). The cumulative emergence follows a sigmoid pattern and I want to create a maximum likelihood estimation of a 3-parameter Weibull cumulative distribution function. The three-parameter models I've been trying to use in the fitdistrplus package keep giving me an error. I think this must have something to do with how my data is structured, but I cannot figure it out. Obviously I want it to read each point as an x (degree days) and a y (emergence) value, but it seems to be unable to read two columns. The main error I'm getting says "Non-numeric argument to mathematical function" or (with slightly different code) "data must be a numeric vector of length greater than 1". Below is my code including added columns in the df_dd_em dataframe for cumulative emergence and percent emergence in case that is useful.
degree_days <- c(998.08,1039.66,1111.29,1165.89,1236.53,1293.71,
1347.66,1387.76,1445.47,1493.44,1553.23,1601.97,
1670.28,1737.29,1791.94,1849.20,1920.91,1967.25,
2036.64,2091.85,2152.89,2199.13,2199.13,2263.09,
2297.94,2352.39,2384.03,2442.44,2541.28,2663.90,
2707.36,2773.82,2816.39,2863.94)
emergence <- c(0,0,0,1,1,0,2,3,17,10,0,0,0,2,0,3,0,0,1,5,0,0,0,0,
0,0,0,0,1,0,0,0,0,0)
cum_em <- cumsum(emergence)
df_dd_em <- data.frame (degree_days, emergence, cum_em)
df_dd_em$percent <- ave(df_dd_em$emergence, FUN = function(df_dd_em) 100*(df_dd_em)/46)
df_dd_em$cum_per <- ave(df_dd_em$cum_em, FUN = function(df_dd_em) 100*(df_dd_em)/46)
x <- pweibull(df_dd_em[c(1,3)],shape=5)
dframe2.mle <- fitdist(x, "weibull",method='mle')
Here's my best guess at what you're after:
Set up data:
dd <- data.frame(degree_days=c(998.08,1039.66,1111.29,1165.89,1236.53,1293.71,
1347.66,1387.76,1445.47,1493.44,1553.23,1601.97,
1670.28,1737.29,1791.94,1849.20,1920.91,1967.25,
2036.64,2091.85,2152.89,2199.13,2199.13,2263.09,
2297.94,2352.39,2384.03,2442.44,2541.28,2663.90,
2707.36,2773.82,2816.39,2863.94),
emergence=c(0,0,0,1,1,0,2,3,17,10,0,0,0,2,0,3,0,0,1,5,0,0,0,0,
0,0,0,0,1,0,0,0,0,0))
dd <- transform(dd,cum_em=cumsum(emergence))
We're actually going to fit to an "interval-censored" distribution (i.e. probability of emergence between successive degree day observations: this version assumes that the first observation refers to observations before the first degree-day observation, you could change it to refer to observations after the last observation).
library(bbmle)
## y*log(p) allowing for 0/0 occurrences:
y_log_p <- function(y,p) ifelse(y==0 & p==0,0,y*log(p))
NLLfun <- function(scale,shape,x=dd$degree_days,y=dd$emergence) {
prob <- pmax(diff(pweibull(c(-Inf,x), ## or (c(x,Inf))
shape=shape,scale=scale)),1e-6)
## multinomial probability
-sum(y_log_p(y,prob))
}
library(bbmle)
I should probably have used something more systematic like the method of moments (i.e. matching the mean and variance of a Weibull distribution with the mean and variance of the data), but I just hacked around a bit to find plausible starting values:
## preliminary look (method of moments would be better)
scvec <- 10^(seq(0,4,length=101))
plot(scvec,sapply(scvec,NLLfun,shape=1))
It's important to use parscale to let R know that the parameters are on very different scales:
startvals <- list(scale=1000,shape=1)
m1 <- mle2(NLLfun,start=startvals,
control=list(parscale=unlist(startvals)))
Now try with a three-parameter Weibull (as originally requested) -- requires only a slight modification of what we already have:
library(FAdist)
NLLfun2 <- function(scale,shape,thres,
x=dd$degree_days,y=dd$emergence) {
prob <- pmax(diff(pweibull3(c(-Inf,x),shape=shape,scale=scale,thres)),
1e-6)
## multinomial probability
-sum(y_log_p(y,prob))
}
startvals2 <- list(scale=1000,shape=1,thres=100)
m2 <- mle2(NLLfun2,start=startvals2,
control=list(parscale=unlist(startvals2)))
Looks like the three-parameter fit is much better:
library(emdbook)
AICtab(m1,m2)
## dAIC df
## m2 0.0 3
## m1 21.7 2
And here's the graphical summary:
with(dd,plot(cum_em~degree_days,cex=3))
with(as.list(coef(m1)),curve(sum(dd$emergence)*
pweibull(x,shape=shape,scale=scale),col=2,
add=TRUE))
with(as.list(coef(m2)),curve(sum(dd$emergence)*
pweibull3(x,shape=shape,
scale=scale,thres=thres),col=4,
add=TRUE))
(could also do this more elegantly with ggplot2 ...)
These don't seem like spectacularly good fits, but they're sane. (You could in principle do a chi-squared goodness-of-fit test based on the expected number of emergences per interval, and accounting for the fact that you've fitted a three-parameter model, although the values might be a bit low ...)
Confidence intervals on the fit are a bit of a nuisance; your choices are (1) bootstrapping; (2) parametric bootstrapping (resample parameters assuming a multivariate normal distribution of the data); (3) delta method.
Using bbmle::mle2 makes it easy to do things like get profile confidence intervals:
confint(m1)
## 2.5 % 97.5 %
## scale 1576.685652 1777.437283
## shape 4.223867 6.318481
dd <- data.frame(degree_days=c(998.08,1039.66,1111.29,1165.89,1236.53,1293.71,
1347.66,1387.76,1445.47,1493.44,1553.23,1601.97,
1670.28,1737.29,1791.94,1849.20,1920.91,1967.25,
2036.64,2091.85,2152.89,2199.13,2199.13,2263.09,
2297.94,2352.39,2384.03,2442.44,2541.28,2663.90,
2707.36,2773.82,2816.39,2863.94),
emergence=c(0,0,0,1,1,0,2,3,17,10,0,0,0,2,0,3,0,0,1,5,0,0,0,0,
0,0,0,0,1,0,0,0,0,0))
dd$cum_em <- cumsum(dd$emergence)
dd$percent <- ave(dd$emergence, FUN = function(dd) 100*(dd)/46)
dd$cum_per <- ave(dd$cum_em, FUN = function(dd) 100*(dd)/46)
dd <- transform(dd)
#start 3 parameter model
library(FAdist)
## y*log(p) allowing for 0/0 occurrences:
y_log_p <- function(y,p) ifelse(y==0 & p==0,0,y*log(p))
NLLfun2 <- function(scale,shape,thres,
x=dd$degree_days,y=dd$percent) {
prob <- pmax(diff(pweibull3(c(-Inf,x),shape=shape,scale=scale,thres)),
1e-6)
## multinomial probability
-sum(y_log_p(y,prob))
}
startvals2 <- list(scale=1000,shape=1,thres=100)
m2 <- mle2(NLLfun2,start=startvals2,
control=list(parscale=unlist(startvals2)))
summary(m2)
#graphical summary
windows(5,5)
with(dd,plot(cum_per~degree_days,cex=3))
with(as.list(coef(m2)),curve(sum(dd$percent)*
pweibull3(x,shape=shape,
scale=scale,thres=thres),col=4,
add=TRUE))
When I run the example code for a random walk using the fractal package:
library(fractal)
## create test series
set.seed(100)
x <- rnorm(1024)
walk <- cumsum(x)
## calculate the Hurst coefficient of a random
## walk series using various techniques
methods <- c("aggabs","aggvar","diffvar","higuchi")
z <- lapply(methods, function(method, walk){
hurstBlock(ifelse1(method=="higuchi",diff(walk),walk), method=method)
},walk=walk )
names(z) <- methods
## plot results
old.plt <- splitplot(2,2,1)
for (i in 1:4){
if (i > 1)
splitplot(2,2,i)
plot(z[[i]], key=FALSE)
mtext(paste(attr(z[[i]],"stat.name"), round(as.numeric(z[[i]]),3), sep=", H="),
line=0.5, adj=1)
}
par(old.plt)
I do not get hurst exponent of .5? I understand the there are different method of calculations but I believe AggVar is most like the R/S rescale method and for a brownian motion this should return .5 not .94 like it does.
SHouldn't it be .5?
Also if you change
walk <- cumsum(x) to
walk <- x Then this is white noise and I should get a Hurst Exponent of 0. But I dont.
Can you help clarify?
I have done extensive testing of the three most prominent Hurst exponent libraries since my answer above and posted a video on youtube, However, from fractal, the most common, Hurst estimator used is the hurstpec from fractal and the hurstexp from the pracma library. The other estimator from PerformanceAnalytics econometrics finance library does not seem to be right no matter what estimate you use. I left on the Authors Github maintenance site an error correction request or check, I have not checked back to see the response, I left the request 4 months ago.
iwm_rt <- dailyReturn(IWM, type ="log")
head(iwm_rt)
```{r}
library(pracma)
library(fractal)
x <- iwm_rt
hurstBlock(x, method="aggAbs", scale.min=8, scale.max=NULL,
scale.ratio=9, weight=function(x) rep(1,length(x)), fit=lm)
fit_aggab <- hurstBlock(x, method="aggAbs", scale.min=8, scale.max=NULL,
scale.ratio=9, weight=function(x) rep(1,length(x)), fit=lm)
plot(fit_aggab)
RoverS(x, n.block.min=22, scale.ratio=9, scale.min=8)
hurstSpec(x, method="standard", freq.max=0.25, dc=FALSE, n.block=NULL,
weight=function(x) rep(1,length(x)), fit=lm, sdf.method="direct")
hurstexp(x)
The aggabs method from hurstblock in fractal is an aggregate estimate from a large sample time series, the most crucial factor in fitting the aggabs estimator is the scale.ratio estimate I solve it by trial and error which is a bit tedious but I do not know of a more scientific method, the scale. ratio is the multiplier for each subsequent series of samples from the time series
I am trying find the probability that a point lies within an ellipse?
For eg if I was plotting the bivariate data (x,y) for 300 datasets in an 95% ellipsoid region, how do I calculate how many times out of 300 will my points fall inside the
ellipse?
Heres the code I am using
library(MASS)
seed<-1234
x<-NULL
k<-1
Sigma2 <- matrix(c(.72,.57,.57,.46),2,2)
Sigma2
rho <- Sigma2[1,2]/sqrt(Sigma2[1,1]*Sigma2[2,2])
rho
eta1<-replicate(300,mvrnorm(k, mu=c(-1.59,-2.44), Sigma2))
library(car)
dataEllipse(eta1[1,],eta1[2,], levels=c(0.05, 0.95))
Thanks for your help.
I don't see why people are jumping on the OP. In context, it's clearly a programming question: it's about getting the empirical frequency of data points within a given ellipse, not a theoretical probability. The OP even posted code and a graph showing what they're trying to obtain.
It may be that they don't fully understand the statistical theory behind a 95% ellipse, but they didn't ask about that. Besides, making plots and calculating frequencies like this is an excellent way of coming to grips with the theory.
Anyway, here's some code that answers the narrowly-defined question of how to count the points within an ellipse obtained via a normal distribution (which is what underlies dataEllipse). The idea is to transform your data to the unit circle via principal components, then get the points within a certain radius of the origin.
within.ellipse <- function(x, y, plot.ellipse=TRUE)
{
if(missing(y) && is.matrix(x) && ncol(x) == 2)
{
y <- x[,2]
x <- x[,1]
}
if(plot.ellipse)
dataEllipse(x, y, levels=0.95)
d <- scale(prcomp(cbind(x, y), scale.=TRUE)$x)
rad <- sqrt(2 * qf(.95, 2, nrow(d) - 1))
mean(sqrt(d[,1]^2 + d[,2]^2) < rad)
}
It was also commented that a 95% data ellipse contains 95% of the data by definition. This is certainly not true, at least for normal-theory ellipses. If your distribution is particularly bad, the coverage frequency may not even converge to the assumed level as the sample size increases. Consider a generalised pareto distribution, for example:
library(evd) # for rgpd
# generalised pareto has no variance for shape > 0.5
z <- sapply(1:1000, function(...) within.ellipse(rgpd(100, shape=5), rgpd(100, shape=5), FALSE))
mean(z)
[[1] 0.97451
z <- sapply(1:1000, function(...) within.ellipse(rgpd(10000, shape=5), rgpd(10000, shape=5), FALSE))
mean(z)
[1] 0.9995808