5.1.5.  VAR_DEPEND

The variable attribute VAR_DEPEND lists all independent variables on which the current variable depends. The number of independent variables listed must be equal to the number of semicolon-separated fields in VAR_SIZE, and the order in which the variables are listed must correspond to the order in which their sizes are given in VAR_SIZE. 

Type: STRING

Entry:  n semicolon-separated fields, with n equal to number of semicolon-separated fields in 

VAR_SIZE

Format: Appropriate subset of the DATA_VARIABLES(cf. Section 4.2.6) entry

Example 1: VAR_DEPEND = DATETIME

Example 2: VAR_DEPEND = DATETIME;LONGITUDE;LATITUDE;ALTITUDE

Example 3: VAR_DEPEND = CONSTANT

Constants must have the following attributes:

VAR_DEPEND = CONSTANT

VAR_SIZE = 1

An axis variable(DATETIME, LATITUDE, LONGITUDE, and e.g. ALTITUDE), which is by definition independent, must use its own variable name as a dependent variable:

VAR_NAME = DATETIME

VAR_DEPEND = DATETIME

VAR_SIZE = n

Independent variables, which are not axis variablesmust have the following attributes:

VAR_DEPEND = INDEPENDENT

VAR_SIZE = n

Multiple occurrencesof a variable in VAR_DEPEND are allowed.

When using older versions of HDF4 (prior to HDF4.2r2) some additional attention is drawn on the length of the name parameter saved with an SDS and its reference through VAR_DEPEND. Please see full discussion in Section 6.1.2.

Four different cases for VAR_DEPEND values for an ALTITUDE variable. Only in the last two cases the ALTITUDEvariable is an axis variable:

All measurements are taken at a constant single height:

VAR_DEPEND = CONSTANT

Measurements are taken at a single height and this height differs over time:

VAR_DEPEND = DATETIME

Measurements have a height axis and the height grid is constant within the file:

VAR_DEPEND = ALTITUDE

Measurements have a height axis and the height grid differs over time:

VAR_DEPEND = DATETIME;ALTITUDE