Point type vs Value type

[jdkent]

how does "type" defined in Point differ from the "type" defined in Value? Do they represent the same concept?
I think @tyarkoni has already explained this to me, but I cannot find my notes on it. @tyarkoni, could you remind me again if you have the time?

Point:

NIMADS/definitions.json

Lines 22 to 29 in 8b40695

	"Point": {
	"properties": [
	["coordinates", "nimads:Coordinates"],
	["space", "schema:Text"],
	["value", "nimads:Value"],
	["type", "schema:Text"]
	]
	},

Value:

NIMADS/definitions.json

Lines 30 to 35 in 8b40695

	"Value": {
	"properties": [
	["type", "schema:Text"],
	["value", "schema:Float"]
	]
	},

[tsalo]

A Point can have multiple values of different types, right? For example, the parameter estimate (beta), variance of the estimate (varcope), test statistic (z, t, F, etc.), and even a standardized effect size (Cohen's d, Hedges g, etc.). I would imagine that information should go in the Value's "type" property, rather than in the Point's.

Perhaps Point "type" should be whether the Point refers to a cluster peak, a cluster sub-peak, or a cluster center of mass?