Graph Dates and Values

Gabriel Genellina gagsl-py2 at yahoo.com.ar
Tue Mar 10 17:44:41 CET 2009


En Tue, 10 Mar 2009 13:32:10 -0200, brianrpsgt1 <brianlong at cox.net>  
escribió:
> On Mar 10, 7:40 am, "Gabriel Genellina" <gagsl-... at yahoo.com.ar>
> wrote:
>> En Tue, 10 Mar 2009 05:08:41 -0200, brianrpsgt1 <brianl... at cox.net>  
>> escribió:
>>
>> > I am trying to plot dates and values on a graph using matplotlib.
>> > Below is the code.  I can run this and it works great, until I get to
>> > about 2000 rows from the DB.  Things really start to slow down.  I
>> > have successfully plotted up to 5000 rows from the DB, but it is very
>> > slow.  I am attempting to plot values for a day, which would be equal
>> > to 84600 records.  Is there a more efficient may to accomplish this?

>> Without looking at the matplotlib docs, the above [] suggests that both  
>> date2num and plt.plot take a list of values to act upon, and you're  
>> feeding one point at a time. Probably you end up creating one series  
>> per  
>> point (instead of a single series with many points). I guess something  
>> like this should work:
>>
>> x, y = zip(*value_data) # "transpose"
>> dates = mdates.date2num(x)
>> plt.plot(dates, y, 'bo', ms=6)
>>
>
> Thanks for the notes.  That is exactly what I thought the problem
> was.  Here is an update.  I put a limit to 100 on the SQL Query to
> test.  When I run your code, I get the data returned, however, I get
> the same return equal to the limit I set.  In other words, when I run
> with a limit of 100, I get the same result 100 times.  Which would
> mean that when I try to run a whole day (86400 :) - it was late!), I
> am getting the same result 86400 times and then it is tyring to plot
> that.
>
> Output below:
>
> [ 733414.06489583  733414.06490741  733414.06491898  733414.06493056 ...
>   733414.06600694  733414.06601852  733414.06603009  733414.06604167]
> (95, 95, 95, 95, ...  95, 95, 95, 94)
>
> If I run this code:
>
> for s in value_data:
>     x = mdates.date2num([s[0]])
>     y = [s[1]]
>     print [x, y]
>
> The results returned are the following:
>
> There are 100 rows in the database
> [ 733414.06489583] [95]
> [ 733414.06490741] [95]
> [ 733414.06491898] [95]
> [ 733414.06493056] [95] ...
> [ 733414.06600694] [95]
> [ 733414.06601852] [95]
> [ 733414.06603009] [95]
> [ 733414.06604167] [94]

Well, both look the same values to me... what's wrong? Why do you say "the  
same results 100 times".

Oh, the code fragment I posted is suposed to *replace* the original for  
loop. Don't put it inside a loop.

-- 
Gabriel Genellina




More information about the Python-list mailing list