tsdata.interpolation
datatypes: tsdata.interpolation
The interpolation method of a timeseries.
A tsdata.interpolation object names how a timeseries is
evaluated between its samples, and holds the function that does it. It is
what ts.DataInfo.Interpolation returns, and it is the value
setinterpmethod sets.
Two methods are built in, 'linear' and 'zoh' (zero-order
hold, which keeps each value until the next sample), and any function
handle may be given instead.
Source Code: tsdata.interpolation
The tsdata.interpolation class contains the following properties:
A function handle called as
newData = fh (newTime, oldTime,
oldData), as in MATLAB: newTime and oldTime are
column vectors, and oldData and newData have time along
their first dimension, a sample per row. It is [] for an object
that names no method. Assigning a function handle sets Name to
'myFuncHandle'.
MATLAB stores this property in several shapes, among them a cell holding the handle and, for a built-in method, a cell holding a handle and a function name; here it is always the handle itself.
One of 'linear', 'zoh' or 'myFuncHandle', or
'' for an object that names no method. Assigning
'linear' or 'zoh' also sets Fhandle to the
built-in function; 'myFuncHandle' is set by assigning a function
handle to Fhandle, not by name. MATLAB accepts any name and
leaves Fhandle as it was, so the two can disagree there.
The tsdata.interpolation class offers the following public methods:
tsdata.interpolation: ip = tsdata.interpolation ()
tsdata.interpolation: ip = tsdata.interpolation (name)
tsdata.interpolation: ip = tsdata.interpolation (fh)
ip = tsdata.interpolation () returns an object that names
no method, with Name '' and Fhandle [].
ip = tsdata.interpolation (name) returns the built-in
method name, 'linear' or 'zoh', given as a
character vector or a string scalar.
ip = tsdata.interpolation (fh) returns a method
evaluated by the function handle fh, called as
newData = fh (newTime, oldTime,
oldData), and names it 'myFuncHandle'.
A cell array is refused, where MATLAB returns an empty object for it without an error.
See also: timeseries, tsdata.datametadata
tsdata.interpolation: S = get (ip)
tsdata.interpolation: value = get (ip, name)
tsdata.interpolation: values = get (ip, names)
S = get (ip) returns a structure of every
property of ip. value = get (ip,
name) returns the property name, a character vector or a
string scalar matched in any case, and values = get
(ip, names) a row cell array of the properties named in
the cell array names. For an array, one name gives a cell
array of its size, and no name, or several, a cell array with a row
per object and a column per property.
See also: tsdata.interpolation.set, timeseries.get
tsdata.interpolation: set (ip, name, value, …)
tsdata.interpolation: ip2 = set (ip, name, value, …)
tsdata.interpolation: S = set (ip)
set (ip, name, value, …) sets each
property name, matched in any case, to its value, in the
order given, in the variable ip of the caller, which must
therefore be a variable. Metadata held by a timeseries is set
with dot assignment instead, as
ts.TimeInfo.Units = 'days'.
ip2 = set (ip, name, value, …)
returns the modified object and leaves ip as it was.
S = set (ip) returns what get (ip)
does.
See also: tsdata.interpolation.get, timeseries.set