The RATS Software Forum

Hi,

I am new in RATS and I am struggling with another task I want to perform:

I am looking to calculate co-skewnes and co-curtosis for a set of series (more than 500), but with little success. I was hoping to get if possible NxN matrix of the values but any solution would be helpful as long as it can be manageable for a big data set..

I checked on the forum but once again could not find a topic that addresses my problem and will deeply appreciate any help with it

Regards

The co-kurtosis of 500 series has something like 2 billion entries. I assume that's not what you want. You'll have to be a bit more specific, since I don't see what is getting you NxN matrices.

Tom,

I now realise that I might be a bit overambitious. What I am after is being able to calculate co-skew and co-curtosis for a number of time series. I guess a table format of (i)xN for multiple should do.

For example co-skew and co-curtosis for (a) against b, c, d etc, then b against a, c, d etc. If I can do that with one operation that would be great

Thanks again for your help

Bobby

That's what gets you 2 billion entries. The co-kurtosis has a value for every combination of 4 indices, allowing for repetitions, from the 500 series.

Tom,

I gave up doing a wholesale calculation (500 series) and was thinking of doing it with 3-4 series. Instead of doing multiple entries I was hoping to get it in a loop.

Thanks