Re: [Discuss-gnuradio] Flowgraph Control

[Richard Bell]

Thank Johannes. Would you be willing to share your python code so I could learn from it?

I'm still trying to get the simulation working with GUIs. I've still got issues getting the GUI window from one run to close properly, without generating errors in terminal. But your approach will help me understand the non-GUI related matters better.

Appreciated,

Rich

On Fri, Oct 23, 2015 at 2:44 AM, Johannes Demel <address@hidden> wrote:

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Richard,

I'm currently working on something quite similar. [1] shows an example
fur a dummy.
Basically what I do: I wrapped my flowgraph into another Python thread
which runs the flowgraph and generates statistics and everything else
I want for my simulation infrastructure.
I fixed some problems in the block 'ber_bf' to shut down the flowgraph
correctly when that limit is exceeded. And not shut down before any
errors occured.
After your flowgraph is done, just get the number of bit errors with
tb.<ber_bf_block_name>.total_errors(). I know it'll break encapsulation.
tb.<ber_bf_block_name>.nitems_read(0) will return the number of bytes.
Be careful with the bit vs. byte issue. But that also depends on your
input data.
My flowgraphs are capped by the FEC decoder, all the other blocks
mostly wait for it. So I just spawn a couple of them in parallel and
delete a flowgraph once it finished and I got all the results.
My assumption is that flowgraph instantiation is really fast compared
to the simulation. Worst case setup time < 2s and simulation time may
exceed 1000s. Also, you don't have to worry about incomplete resets
for your flowgraph.

I hope I could show a different approach here.

Also, I use 'no GUI' flowgraphs.

Cheers
Johannes


[1] http://imgur.com/wDOPqHL

On 22.10.2015 23:26, Richard Bell wrote:
> I removed the qt app.exec_() line and that seemed to fix it
>
> On Thu, Oct 22, 2015 at 2:16 PM, Richard Bell
> <address@hidden> wrote:
>
>> Thanks.
>>
>> I'm running into a problem getting tb.stop() to trigger. I tried
>> putting a head block in the flow graph thinking it triggers
>> WORK_DONE which I thought would lead to tb.stop(), but it
>> doesn't. The flowgraph window just sits there frozen. How do I
>> get a running flow graph to trigger tb.stop(), so the next
>> iteration can start?
>>
>> Rich
>>
>> On Thu, Oct 22, 2015 at 1:21 PM, West, Nathan
>> <address@hidden
>>> wrote:
>>
>>>
>>>
>>> On Thu, Oct 22, 2015 at 3:58 PM, Richard Bell
>>> <address@hidden> wrote:
>>>
>>>> Or is this the correct stategy:
>>>>
>>>> if __name__ == '__main__':
>>>>
>>>> tb1 = my_flowgraph( paramter_set_1 ) tb.start() tb.wait()
>>>> tb.stop()
>>>>
>>>> tb2 = my_flowgraph( paramter_set_2 ) tb.start() tb.wait()
>>>> tb.stop()
>>>>
>>>> tb3 = my_flowgraph( paramter_set_3 ) tb.start() tb.wait()
>>>> tb.stop()
>>>>
>>>> etc..
>>>>
>>>
>>> That's what I would do, except create a list of your parameter
>>> values and loop over the list.
>>>
>>>
>>
>
>
>

> _______________________________________________ Discuss-gnuradio
> mailing list address@hidden
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
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=ez4I
-----END PGP SIGNATURE-----

_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio