From MdsWiki
MDSplus supports numerous data types, from simple numeric or text types to composite entities which combine related information into a structure. Primitive data types include signed and unsigned integers of 1, 2, 4, 8 and 16 bytes; single and double precision floats; single and double precision complex numbers; and character data. Arrays of these data types with up to 7 dimensions are supported. Some of the composite data types are used internally to represent, for example, data acquisition devices or compiled representations of expressions and data acquisition actions, and are not ordinarily accessed directly by users. Users will normally be accessing the primitive data types and several of the composite MDSplus data types such as SIGNALS, DIMENSIONS, WITH UNITS, WITH ERRORS and RANGES. The datatype of a given node can be determined by evaluating the expression GETNCI(node_reference,"characteristic").
One of the more important data types provided by MDSplus is the signal. A signal is a structure which combines a value part, optionally a raw data part, and one or more dimension descriptions. Data from measurement devices as well as analysis results are often stored as signals. Many visualization applications will generate plots of signals by plotting the value portion versus the dimensions. The value part of a signal can be an expression which references the raw data part of the signal. Support for transient recorders often use this feature by storing the raw counts values as the raw data part and use an expression to convert this raw data to engineering units (such as voltage) in the value part of the signal. In the case of the support for a transient record, a representation of the time base would be stored in the first dimension part of the signal. The components of a signal can be accessed by evaluating the signal with the TDI functions DIM_OF, DATA, DATA_WITH_UNITS, RAW_OF.
More details on datatypes can be found in the users manual; see DATATYPES