Convert Distributed Array to Matrix

I have a distributed array created using:

cal_data=distributed.zeros(1,2);

I would like to convert the array to a matrix, to use the curve fitting toolbox:

xdata=zeros(1,2);
xdata(:)=cal_data;

I get the following error:

The following error occurred converting from distributed to double: Conversion to double from distributed is not possible

Please help me.

Thanks for your time,

Jake Brinkmann

R2011b

Answer 1

[Jill Reese]

The functionality you are looking for is "gather":

xdata = gather(cal_data);

If you don't mind my asking, what particular functionality are you trying to use in the Curve Fitting Toolbox? We are always interested to hear what operations our customers want to use with distributed arrays.

[Jacob Brinkmann]

>>[fresult,gof,output]=fit(xdata,ydata,'poly1',opts);
Warning: Converting X to matrix of double. 
> In fit>iCreateWarningFunction/nThrowWarning at 738
  In fit>iFit at 144
  In fit at 109
  In tirs_datasets at 104
  In tirs_tool_v2>files_calculate_button_Callback at 274
  In gui_mainfcn at 96
  In tirs_tool_v2 at 50
  In @(hObject,eventdata)tirs_tool_v2('files_calculate_button_Callback',hObject,eventdata,guidata(hObject)) 
Error using distributed/subsref (line 33)
Distributed SUBSREF currently only supports () indexing.

Error in fit>iFit (line 461)
                [~, A] = feval(model,coefftemp{:},probparams{:},xdatacell{:});

Error in fit (line 109)
[fitobj, goodness, output, convmsg] = iFit( xdatain, ydatain, fittypeobj, ...

[Jill Reese]

Jacob,

Unfortunately, some of the MATLAB functions used by the Curve Fitting Toolbox's fit function are not currently implemented for distributed arrays. In the interim, if you need a simple linear fit to data you might check if polyfit has the functionality you need. I tried some simple examples and it seems like polyfit works with distributed arrays if the first input argument is not distributed but the second one is.

B=1:1000;
A=2*B-4;
z=polyfit(B,distributed(A),1);

Answer 2

[shrouk mohamed]

Thank you, this was helpful.