# How do I interpret this Weibull plot?

I was exploring Weibull analysis for understanding reliability of two specific specimen. I used the R package, "weibulltoolkit" and "survival" to get the plot in question. The dataset is big so I am not posting it.

d <- data.frame(ob=X, state=1)
s = Surv(d$ob, d$state)
plot.wb(s)

From my understanding, both plots indicate bad fits! This could be an artifact of having too many data-points in the system or could be due to seasonal-effects or other unknown effects. Looking at the plot, what am I supposed to understand? Specifically,

• Plot 1 (using 4000 points) seems to indicate the presence of two different behaviors (0.05 - 5 and 5-1e+05). Does this mean that I am better off splitting the dataset into two and analyzing them separately? If so, what is the implication of this?
• Same seems to be going for Plot 2 (using 400 points) i.e. I would be better of splitting this dataset as well.

What is a good way of comparing these two datasets? Any suggestions?

EDIT: Upon Googling, I came across something called Multiple failure modes though I am not sure how to do this in R. Any suggestions would be greatly appreciated

### PLOT 2

-
What is your goal? Do you really need Weibull, or any other parametric analysis? The survival package has loads of nonparametric methods. –  Aniko Aug 23 '11 at 14:27
Well, my goal is to compare the reliability of these two specimen. If there is a better way to do this, please feel free to suggest. It is just that I came across Weibull yesterday and thought of playing around with it. –  Legend Aug 23 '11 at 18:54
@Legend, work on that accept rate! :) –  Brandon Bertelsen Aug 24 '11 at 2:44
@Brandon Bertelsen: Done! Hope it looks better now :) –  Legend Aug 24 '11 at 3:34