Re: 3.4.2 MinGW: bode.m (control-2.1.55) gives error

[Lukas Reichlin]

On 19.09.2011, at 15:10, Bernhard Weller wrote:

> 
> Lukas Reichlin-2 wrote:
>> 
>> 
>> Hi Bernhard
>> 
>> On my Mac running Octave 3.4.2 I can't reproduce your error. The attached
>> m-file runs without a single warning or error and produces the plot. It
>> would be good if some other people - especially on Windows - could try out
>> the script and see if it works for them.
>> 
>> Would it help if you change lines 39 and 40 in @lti/tfdata.m from
>> 
>>    num = cellfun ("@tfpoly/get", num, "uniformoutput", false);
>>    den = cellfun ("@tfpoly/get", den, "uniformoutput", false);
>> 
>> to
>> 
>>    num = cellfun (@get, num, "uniformoutput", false);
>>    den = cellfun (@get, den, "uniformoutput", false);
>> 
>> ? Maybe you need to restart Octave after you made these changes.
>> 
>> Regards,
>> Lukas
>> 
>> 
> 
> Hey Lukas,
> 
> thanks for the fix, now it works like a charm, no warnings, no errors.
> 
> I just made those changes and restarted Octave.
> 
> But it would be still interesting to know if this is a Windows related error
> or whether something went wrong in my installation.
> 
> Thanks,
> Bernhard

Hi Bernhard

I made the same changes to @tf/__get__.m. My plan is to release control-2.2.0 
within two weeks or so and incorporate those fixes together with other 
improvements. Most notably MIMO-capable transfer function to state-space 
conversion.

AFAIK 3.4.2 MinGW is still considered to be somewhat experimental, but the 
"@tfpoly/get" statement was unecessary and bad style for sure.

Regards,
Lukas