Missing values

[Tom Richardson]

Dear Group,

I am very interested in the possibility of using transfer entropy to identify leader-follower relationships between multiple pairs of time-series.

However, my time-series have many missing values in them, and the timing of these missing values is not synchronized across individual time-series. 

Further, JIDT GUI refuses to load the data, delivering a   'For input string: "NA" ' popup message.

So, before diving in, I was wondering whether someone could please clarify whether the command-line versions have the facility to deal with missing values.. ?

Best Regards,

Tom

[Joseph Lizier]

Hi Tom,

The short answer is yes. There are two things you need to do:

1. The invalid values still need to be somehow readable as numbers in your input file, even though we won't be using them later. So I'd suggest you open the text file, and do a find and replace on the NA values as 0 or something else, then save the files with another name. You should then be able to load that file into the GUI -- check that this works ok. That won't give you the right answer yet though.

2. Next, you will need to create a boolean array for your each of your source and target time series which indicates whether each point in the time series is valid (true) or missing (false). You could do that by looking for those NA values in your original file somehow. You then call the setObservations(double[] source, double[] destination, boolean[] sourceValid, boolean[] destValid) method with all of these. The estimator will work out which tuples it is able to use or not from the samples then.

A couple of caveats: that's not available on every estimator (it is there for KSG), and I don't think it works with auto-embedding (that should be eminently fixable, but not something I have time for atm).

I hope that helps,

--joe
+61 408 186 901 (Au mobile)

--
JIDT homepage: http://code.google.com/p/information-dynamics-toolkit/
---
You received this message because you are subscribed to the Google Groups "Java Information Dynamics Toolkit (JIDT) discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jidt-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to jidt-d...@googlegroups.com.
Visit this group at https://groups.google.com/group/jidt-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/jidt-discuss/cb4e6edf-a2eb-4414-9e3a-09fd68a6abb1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.