<div dir="auto"><div>Thank you Dr. Ben. I will try to following this slowly and match my evolve.py file with the new one you've provided :)<div dir="auto"><br></div><div dir="auto">Cheers, Yusak</div><br><div class="gmail_extra"><br><div class="gmail_quote">On Sep 24, 2017 8:28 PM, "Ben Elliston" <<a href="mailto:bje@air.net.au">bje@air.net.au</a>> wrote:<br type="attribution"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Yusak<br>
<div class="quoted-text"><br>
On Thu, Sep 21, 2017 at 03:35:00PM +1000, Ben Elliston wrote:<br>
<br>
> When you run replay.py, you need to supply the same -d options as you<br>
> pass to evolve.py. I have some ideas for how to make this better, but<br>
> haven't implemented them yet.<br>
<br>
</div>Thanks for poking me about this last week. This shortcoming annoyed<br>
me, but not enough to fix it. Now that others are suffering, I fixed<br>
it. :-)<br>
<br>
Check the notebook documentation for details, but in summary, the<br>
evolve.py script now writes out everything that was used as part of<br>
the evolution so that replay.py can replicate the results without you<br>
having to supply that information. The input parameters, generator<br>
capacites, scenario name, etc. are all encoded in a JSON format file.<br>
You then just pass the filename to replay.py like so:<br>
<br>
python replay.py -f results.json [plus any other options like -x]<br>
<br>
Enjoy!<br>
<font color="#888888">Ben<br>
</font></blockquote></div><br></div></div></div>