Page 1 of 1

Cross Sectional Percentile Ranking Across Time Series

Posted: Mon Feb 10, 2014 6:41 am
by telaxin
Hi,

I was wondering whether it would be possible and how do I go about doing a percentile rank across various time series for the same date. Thanks!

Re: Cross Sectional Percentile Ranking Across Time Series

Posted: Mon Feb 10, 2014 9:06 am
by EViews Gareth
Could you define what you mean a little more explicitly?

Re: Cross Sectional Percentile Ranking Across Time Series

Posted: Mon Feb 10, 2014 4:24 pm
by telaxin
For example:
TS1 TS2 TS3
1/1/2000 5 4 3


There are 3 time series TS1 TS2 and TS3. I want to do a percentile ranking row wise such that:

TS1 TS2 TS3
1/1/2000 100% 50% 0%

So I do a percentile ranking across rows across multiple timeseries

Re: Cross Sectional Percentile Ranking Across Time Series

Posted: Mon Feb 10, 2014 4:53 pm
by EViews Gareth
There isn't anything built in that will do that for you. All you can do is convert a group containing the series into a matrix object, pull off each row of the matrix one at a time into a vector object, use the @ranks function to get you the relative ranking of each row, then calculate the percentiles.