@kyle very cool! We would be early users of that. In fact, I plan to watch
your work for future adoption -- let me know if I can help test or develop.
In the meantime, I came up with this rather hacky attempt. Should work for
us in the mean time. Posting a sketch here in case it's of use/interest to
anyone:
from IPython.core.magic import register_line_magic
from IPython.display import display, Javascript
@register_line_magic
def rerun(line):
if len(line) > 0:
try:
seconds = int(line)
except:
raise ValueError('Could not understand %d as a time in
(integral) '
'seconds after which to rerun the cell')
else:
seconds = 10
js = """
var interval = %d;
var cell_elt = this.element[0].parentElement.parentElement;
var cell = $(cell_elt).data('cell');
var idx = IPython.notebook.find_cell_index(cell);
var lock_name = "rerun_lock_" + idx
if (sessionStorage.getItem(lock_name) === null) {
sessionStorage.setItem(lock_name, "off");
console.log('setting first lock for cell :: ' + idx + ' :: OFF');
};
element.append('cell will rerun in: ' + interval + ' seconds');
if ( sessionStorage.getItem(lock_name) === 'off' ) {
console.log('rerunning cell :: ' + idx);
sessionStorage.setItem(lock_name, "on");
console.log('lock for cell :: ' + idx + ' :: ON');
setTimeout(function(){
cell.execute();
console.log('ran cell ' + idx);
sessionStorage.setItem(lock_name, "off");
console.log('lock for cell :: ' + idx + ' :: OFF');
}, interval * 1000);
};
""" % seconds
js_obj = Javascript(js)
display(js_obj)
return line
On Friday, February 3, 2017 at 2:01:21 PM UTC-8, rgbkrk wrote:
>
> Hey TJ,
>
> I'm looking forward to the update display mechanics that are coming to a
> later release - see https://github.com/ipython/ipython/pull/10048
>
> I created an animation below that updates a figure in place (full
> replacement) every tenth of a second. This could also be done on a thread
> pool, targeting just one display:
>
>
>
>
> -- Kyle
>
>
> On Wed, Feb 1, 2017 at 7:02 PM, TJ Lane <[email protected]
> <javascript:>> wrote:
>
>> Hi Jupyter Community,
>>
>> I am interested in setting up some very simple, low-performance "live"
>> plots using matplotlib. Think line plot updating at ~1 Hz. Simplicity and
>> keeping the "static" matplotlib interface intact are priorities.
>>
>> My current vision for this is to write a magic function, %rerun <time>.
>> This would declare that the code in that cell should be re-run after a
>> specified time, perhaps defaulting to 1 or 10 seconds. The notebook could
>> block while the cell is being re-evaluated, but should not block while
>> waiting. This would let me write something like:
>>
>> >>> %rerun
>> >>> data = fetch_new_data() # gets data from external server
>> >>> plt.plot(data)
>> >>> plt.show()
>>
>> and get a new plot every second or so.
>>
>> I am wondering if someone can give me some advice on how best to
>> implement this in the notebook framework... I am a long time user, but
>> unfamiliar with the architecture "under the hood". If people can anticipate
>> issues (or explain why this is a terrible idea) that would be most
>> appreciated.
>>
>> I am currently thinking of trying to catch the end of the cell execution,
>> and at that time, spawning a thread. This thread would just sleep for the
>> specified time, before waking up, calling for the (re)execution of the
>> specified cell, and then dying. Calling the cell again would spawn a new
>> thread that would re-start the cycle. At any point, by removing or
>> commenting the %rerun line, the user could break the cycle.
>>
>> That plan brings up a number of specific questions:
>> * how can I find out what cell code is being executed in?
>> * how can I programmatically execute code in a cell?
>> * are there issues with spawning threads from the nb kernel?
>>
>> I have searched around a bit and not found much documentation or many
>> examples that made sense to my uninitiated mind... so thanks much for any
>> help.
>>
>> Best,
>>
>> TJ
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "Project Jupyter" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected] <javascript:>.
>> To post to this group, send email to [email protected]
>> <javascript:>.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/jupyter/e85c1e1d-37e0-46af-825d-cda214d1e930%40googlegroups.com
>>
>> <https://groups.google.com/d/msgid/jupyter/e85c1e1d-37e0-46af-825d-cda214d1e930%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>
>
> --
> Kyle Kelley (@rgbkrk <https://twitter.com/rgbkrk>; lambdaops.com)
>
--
You received this message because you are subscribed to the Google Groups
"Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/bfe4e1df-6ab1-41eb-bdf2-1bd3ed20a41a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
