> 
> lc.mod1 <- nls(rel_eff ~ g0 + g1*loc + g3*exp(g2*week), 
+            start=c(g0=1.025, g1=-0.0459, g2=-0.122, g3=-0.5))
> 
> summary(lc.mod1)

Formula: rel_eff ~ g0 + g1 * loc + g3 * exp(g2 * week)

Parameters:
    Estimate Std. Error t value Pr(>|t|)    
g0  1.015598   0.003672  276.60  < 2e-16 ***
g1 -0.047267   0.004109  -11.50 1.06e-11 ***
g2 -0.134795   0.004360  -30.92  < 2e-16 ***
g3 -0.552442   0.008157  -67.72  < 2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Residual standard error: 0.01125 on 26 degrees of freedom

Number of iterations to convergence: 4 
Achieved convergence tolerance: 6.967e-06 

> 
> plot(week[loc==1],rel_eff[loc==1],pch=1,xlab="week",ylab="relative efficiency")
> points(week[loc==0],rel_eff[loc==0],pch=2)
> 
> w <- 0:90
> 
> lines(w,predict(lc.mod1,list(loc=1,week=w)),lty=1)
> lines(w,predict(lc.mod1,list(loc=0,week=w)),lty=2)
>