lomb_scargle: Lomb Scargle method

In multiband: Period Estimation for Multiple Bands

Description

lomb_scargle applies the Lomb Scargle method for performing weighted nonlinear least squares in a single band using magnitude measurements recorded at irregular intervals.

Usage

lomb_scargle(t, m, sigma, omega, weights_flag = TRUE)

Arguments

t	times
m	magnitudes
sigma	errors
omega	frequency
weights_flag	boolean (TRUE to use inverse errors as weights; FALSE uses uniform weights)

Examples

n <- 1e3
set.seed(12345)
sigma <- runif(n)
t <- cumsum(runif(n))
A <- 2.3
rho <- 0.1
omega <- 0.2
beta0 <- 0.3
m <- beta0 + A*sin(omega*t + rho) + sigma*rnorm(n)
plot(t,m,xlab='time',ylab='magnitude',main='Simulated Light Curve',pch=16)

## Try several omega
nOmega <- 500
omega_seq <- seq(0.1,0.3,length=nOmega)
sol_ls <- vector(mode="list",length=nOmega)
RSS_seq <- double(nOmega)
for (i in 1:nOmega) {
  sol_ls[[i]] <- lomb_scargle(t,m,sigma,omega_seq[i])
  RSS_seq[i] <- sol_ls[[i]]$RSS
}

plot(omega_seq,RSS_seq,xlab=expression(omega),ylab='RSS',pch=16)
ix_min <- which(RSS_seq==min(RSS_seq))
sol_final <- sol_ls[[ix_min]]

Example output

multiband documentation built on May 2, 2019, 3:30 a.m.