I am interested on opinions on how to treat log tranformed data. I am
working with water quality data - trying to characterize the regions
nationwide by some single central values. I use geometric mean,
calculated from log data as an "average" value. With log data there are
however problems such as the following:
1. What is the equivalent of standard deviation on log transformed
dataset (especially when it comes to back-transforming)? What I did was
to calculate the "normalized standard deviation", which was SD of log
data divided by mean of log data. However I do need some "spread"
parameter in absolute numbers for a general public.The workaround by
using range defined between two back-transformed SD (derived from log
data) seems to be too speculative and non-intuitive to me.
2. While log transforming data, values less than 1 return negative
values. How to deal with this situation?
Thanks for useful replies,
*To post a message to the list, send it to ai-geostats@...
*As a general service to list users, please remember to post a summary
of any useful responses to your questions.
*To unsubscribe, send email to majordomo@...
with no subject and
"unsubscribe ai-geostats" in the message body.
DO NOT SEND Subscribe/Unsubscribe requests to the list!