Details

ValueDimensions is typically used to disambiguate between multidimensional values and lists of scalar values.
Possible settings include:
Automatic automatically find dimensions

1 scalar values

n vector values of length n

{n₁,n₂,…} array values of dimension {n₁,n₂,…}
The n_i must be positive integers.

Examples

Take a matrix of characters:

Wolfram Language code: data = Characters /@ {"dog", "cat", "fox", "owl"}

Create a TemporalData object containing a collection of four scalar-valued paths:

Wolfram Language code: TemporalData[data, Automatic, ValueDimensions -> 1]["Paths"]//Column

Interpret the same data as a single path of dimension 3:

Wolfram Language code: TemporalData[data, Automatic, ValueDimensions -> 3]["Paths"]//Column

Use the data as the value for a single timestamp in a single path of dimensions {4,3}:

Wolfram Language code: TemporalData[{data}, Automatic, ValueDimensions -> {4, 3}]["Paths"]//Column

Top

	Automatic	automatically find dimensions
	1	scalar values
	n	vector values of length n
	{n₁,n₂,…}	array values of dimension {n₁,n₂,…}