Would anyone be able to help me out with the following (or at the least a reference covering the technique (using R) that I require)?; I have a list of events (A, B, C, D), their start date/time and their duration. I also have a time series of a level which is generally maintained at around 90% however it drops at certain times in the year.>From the time series I have extracted the times that the levels drop to 5%and the time until it returns to 20% or higher, so for each drop event, I have the drop start time and drop duration. I would like to determine the following: 1. Probability of an event A-D causing a drop event (i.e., for the the level to drop to less than say 5%) 2. Lag between an event A-D and a drop event (This is less important but would be nice) 3. Statistics describing the drop duration dependent on each event A-D (The average would be a start, however a frequency distribution for each event type would be ideal). Any help whatsoever would be greatly appreached. Thanks, Marie [[alternative HTML version deleted]]