Hi ankita,
On Wednesday 04 November 2009 06:04:55 ankita dutta wrote:
> hi all,
>
> I am working on some graph stuffs and stuck at a point.
>
> I am trying to plot a histogram using simple :
>
> *import matplotlib.pyplot as plt
> .
> .
> plt.hist(x,bins=10,histtype='bar')
plt.show()
> *
> but i want to know
>
> 1) what is the value of a particular bar in histogram and also want to
> print in on output image along with the bars ( either at top of bar or at
> bottom)*
n, bins, patches = plt.hist(...)
gives you the binegdes 'bins' the corresponding frequency / number of values
per bin 'n' and the mplobjects of the shown bars
# e.g. for top of the bar:
n, bins, patches = plt.hist(x0, bins=10, histtype='bar')
for index in xrange(len(bins)1):
plt.text(bins[index], n[index], " %i " % (n[index]))
> *2) How to give input of 2 types of dataset in single hist() command , so
> that i can compare 2 different dataset side by side on my output image. ( i
> dont want it on top of each other as given by the command *histtype*:
> [ ‘barstacked’] and also i read some example in matplotlib site, but
> they are for random numbers : something like this:
>
> x0 = mu + sigma*P.randn(10000)
> x1 = mu + sigma*P.randn(7000)
> x2 = mu + sigma*P.randn(3000)
> P.figure()
> n, bins, patches = P.hist( [x0,x1,x2], 10, histtype='bar')
>
> but i need this comparision for some dataset.
I can't see any problems with placing dataset instead of x0 and x1 into this
call.
Please notice: in that case n is a tuple and each tupleelement corresponds to
the number of values per bin of one dataset.
Kind regards Matthias
