calculate TE from 2D pose

[cms]

i want to calculate TE between the two vehicles.

vehicle's data is 2D pose (x,y) (ex. [0,0], [1,1], [2,2]...)

so i calculated TE using TransferEntropyCalculatorMultiVariateKraskov class by PYTHON

and i set property as follows:

teCalc.setProperty(teCalcClass.PROP_K_SEARCH_MAX, "6")
teCalc.setProperty(teCalcClass.PROP_TAU_SEARCH_MAX, "6")

teCalc.initialise(2,2)

i have some question

1. Is this thr right way?

2. I used this data

 but TE is 0.0 both car1 -> car2 and car2 -> car1

 Am i doing it the wrong way?

3. I made sample data and tested it. As a result, i get a negative value

what does negative value mean?

I'm still a student so i'm lacking. 

I'd appreciate your help.  

[Joseph Lizier]

Hello,

Answers are inline below

--joe

On Fri, 11 Dec 2020 at 09:17, 'cms' via Java Information Dynamics Toolkit (JIDT) discussion <jidt-d...@googlegroups.com> wrote:

i want to calculate TE between the two vehicles.

vehicle's data is 2D pose (x,y) (ex. [0,0], [1,1], [2,2]...)

so i calculated TE using TransferEntropyCalculatorMultiVariateKraskov class by PYTHON

and i set property as follows:

teCalc.setProperty(teCalcClass.PROP_K_SEARCH_MAX, "6")
teCalc.setProperty(teCalcClass.PROP_TAU_SEARCH_MAX, "6")

teCalc.initialise(2,2)

i have some question

1. Is this thr right way?

If you are setting K_SEARCH_MAX and TAU_SEARCH_MAX, this indicates that you are wanted to have these auto-selected (as described in this video for the univariate TE). You will also need to indicate the auto-embedding method (i.e. setting the teCalcClass.PROP_AUTO_EMBED_METHOD property to e.g. teCalcClass.AUTO_EMBED_METHOD_MAX_CORR_AIS_DEST_ONLY) otherwise the embedding parameters will stay at their default of 1.

The initialise(2,2) call looks fine for 2D data.

2. I used this data

 but TE is 0.0 both car1 -> car2 and car2 -> car1

 Am i doing it the wrong way?

I've no idea which data you are talking about, so that is difficult to answer. It is unlikely the KSG estimator gives an exact 0 answer though, so something is likely not quite right.

3. I made sample data and tested it. As a result, i get a negative value

what does negative value mean?

See this FAQ

 

I'm still a student so i'm lacking. 

I'd appreciate your help.  

--
JIDT homepage: https://github.com/jlizier/jidt
---
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...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jidt-discuss/3e15224b-579d-4bba-b80c-0ff35295ee0en%40googlegroups.com.