04-17-2016 11:09 AM
I have summary data.
i need to caclulate median and minimum value for those summary data.
I have other statistics values - standard deviation, variance, average.
so, how we can caclulate median and minimum value using this statistics values ?
let me know, if any questions.
04-17-2016 12:08 PM - edited 04-17-2016 02:51 PM
So, you don't have access to the original data and the only available information is standard deviation, variance (redundant) and average?
This would render the calculation impossible, except for the trivial case that variance=0. With a little more information such as the number of non-missing values (N) and a distributional assumption about the data, at best, you could obtain (possibly very rough) estimates.
Also, how does your post relate to text mining (see forum title)? [Edit: In the meantime an administrator or super user has moved the thread to a more appropriate forum. Thanks!]
04-17-2016 02:53 PM
04-17-2016 03:48 PM
Here is an example:
data test; input x @@; dataset=ceil(_n_/5); cards; 1 10 12 13 14 6 7 8 10 19 ; proc print data=test; run; proc means data=test var std mean median min; by dataset; run;
The two "datasets" (combined into one dataset TEST just to simplify the code) have 5 observations each. They have the same variance, standard deviation and mean (average), but their medians (12 vs. 8) and minimum values (1 vs. 6) differ considerably.