Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #76315 > unrolled thread

Matplotlib Contour Plots

Back to article view | Back to comp.lang.python

Contents

  Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-14 08:22 -0700
    Re: Matplotlib Contour Plots Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-15 02:53 +1000
      Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-15 04:54 -0700
        Re: Matplotlib Contour Plots Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-15 23:23 +1000
          Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-15 07:42 -0700
            Re: Matplotlib Contour Plots Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-16 01:13 +1000
              Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-18 09:51 -0700
                Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-18 09:55 -0700
                  Re: Matplotlib Contour Plots Rustom Mody <rustompmody@gmail.com> - 2014-08-18 10:16 -0700
                  Re: Matplotlib Contour Plots Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-19 10:10 +1000
                    Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-19 01:43 -0700
                      Re: Matplotlib Contour Plots Dennis Lee Bieber <wlfraed@ix.netcom.com> - 2014-08-19 09:12 -0400
                      Re: Matplotlib Contour Plots pecore@pascolo.net - 2014-08-19 23:21 +0200
                        Re: Matplotlib Contour Plots Jamie Mitchell <jamiemitchell1604@gmail.com> - 2014-08-20 03:02 -0700
                          Re: Matplotlib Contour Plots Mark Lawrence <breamoreboy@yahoo.co.uk> - 2014-08-20 17:55 +0100
                Re: Matplotlib Contour Plots Christian Gollwitzer <auriocus@gmx.de> - 2014-08-18 19:49 +0200
            Re: Matplotlib Contour Plots Anssi Saari <as@sci.fi> - 2014-08-18 13:37 +0300

#76315 — Matplotlib Contour Plots

Hello all,

I want to contour a scatter plot but I don't know how.

Can anyone help me out?

Cheers,

Jamie

[toc] | [next] | [standalone]

#76321

Jamie Mitchell wrote:

> Hello all,
> 
> I want to contour a scatter plot but I don't know how.
> 
> Can anyone help me out?

Certainly. Which way did you come in? 

:-)

Sorry, I couldn't resist.

It took me literally 20 seconds to find this by googling for "matplotlib
contour plot", and it only took that long because I misspelled "contour"
the first time.

http://matplotlib.org/examples/pylab_examples/contour_demo.html


Does this help? If not, please explain what experience you have with
matplotlib, what you have tried, what you expected it to do, and what it
did instead.



-- 
Steven

[toc] | [prev] | [next] | [standalone]

#76365

On Thursday, August 14, 2014 5:53:09 PM UTC+1, Steven D'Aprano wrote:
> Jamie Mitchell wrote:
> 
> 
> 
> > Hello all,
> 
> > 
> 
> > I want to contour a scatter plot but I don't know how.
> 
> > 
> 
> > Can anyone help me out?
> 
> 
> 
> Certainly. Which way did you come in? 
> 
> 
> 
> :-)
> 
> 
> 
> Sorry, I couldn't resist.
> 
> 
> 
> It took me literally 20 seconds to find this by googling for "matplotlib
> 
> contour plot", and it only took that long because I misspelled "contour"
> 
> the first time.
> 
> 
> 
> http://matplotlib.org/examples/pylab_examples/contour_demo.html
> 
> 
> 
> 
> 
> Does this help? If not, please explain what experience you have with
> 
> matplotlib, what you have tried, what you expected it to do, and what it
> 
> did instead.
> 
> 
> 
> 
> 
> 
> 
> -- 
> 
> Steven

Yep I've seen that thanks but I can't get it to work. I don't have much experience with matplotlib or programming in general.

I just want to get a contour plot of two numpy arrays.

When I call plt.contour on my data I get "input must be a 2D array"

An example of one of my arrays:

array([ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
        2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
        2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
        1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
        2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
        2.16600013,  2.00999999,  1.86200011,  2.19800019,  2.01200008], dtype=float32)

How do I get the above array in to the right format for a contour plot?

Thanks,

Jamie

[toc] | [prev] | [next] | [standalone]

#76367

Jamie Mitchell wrote:

[...]
> I just want to get a contour plot of two numpy arrays.
> When I call plt.contour on my data I get "input must be a 2D array"

You are providing a 1D array, or possibly a 3D array. So the question you
really want to ask is not "How do I do contour plots" but "how do I make a
2D array?"


> An example of one of my arrays:
> 
> array([ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
>         2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
>         2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
>         1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
>         2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
>         2.16600013,  2.00999999,  1.86200011,  2.19800019,  2.01200008],
>         dtype=float32)
> 
> How do I get the above array in to the right format for a contour plot?

Here's an example of making a 2D array:

py> import numpy
py> a = numpy.array([1.2, 2.5, 3.7, 4.8])  # One dimensional array
py> a
array([ 1.2,  2.5,  3.7,  4.8])
py> b = numpy.array([ [1.2, 2.5, 3.7, 4.8], 
...                   [9.5, 8.1, 7.0, 6.2] ])  # Two dimensional array
py> b
array([[ 1.2,  2.5,  3.7,  4.8],
       [ 9.5,  8.1,  7. ,  6.2]])

One dimensional arrays are made from a single list of numbers: [...]

Two dimensional arrays are made from a list of lists: [ [...], [...] ]



-- 
Steven

[toc] | [prev] | [next] | [standalone]

#76369

On Friday, August 15, 2014 2:23:25 PM UTC+1, Steven D'Aprano wrote:
> Jamie Mitchell wrote:
> 
> 
> 
> [...]
> 
> > I just want to get a contour plot of two numpy arrays.
> 
> > When I call plt.contour on my data I get "input must be a 2D array"
> 
> 
> 
> You are providing a 1D array, or possibly a 3D array. So the question you
> 
> really want to ask is not "How do I do contour plots" but "how do I make a
> 
> 2D array?"
> 
> 
> 
> 
> 
> > An example of one of my arrays:
> 
> > 
> 
> > array([ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
> 
> >         2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
> 
> >         2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
> 
> >         1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
> 
> >         2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
> 
> >         2.16600013,  2.00999999,  1.86200011,  2.19800019,  2.01200008],
> 
> >         dtype=float32)
> 
> > 
> 
> > How do I get the above array in to the right format for a contour plot?
> 
> 
> 
> Here's an example of making a 2D array:
> 
> 
> 
> py> import numpy
> 
> py> a = numpy.array([1.2, 2.5, 3.7, 4.8])  # One dimensional array
> 
> py> a
> 
> array([ 1.2,  2.5,  3.7,  4.8])
> 
> py> b = numpy.array([ [1.2, 2.5, 3.7, 4.8], 
> 
> ...                   [9.5, 8.1, 7.0, 6.2] ])  # Two dimensional array
> 
> py> b
> 
> array([[ 1.2,  2.5,  3.7,  4.8],
> 
>        [ 9.5,  8.1,  7. ,  6.2]])
> 
> 
> 
> One dimensional arrays are made from a single list of numbers: [...]
> 
> 
> 
> Two dimensional arrays are made from a list of lists: [ [...], [...] ]
> 
> 
> 
> 
> 
> 
> 
> -- 
> 
> Steven

Thank you Steven.

I created the 2D array which read as:

array([[[ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
          2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
          2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
          1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
          2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
          2.16600013,  2.00999999,  1.86200011,  2.19800019,  2.01200008]],

       [[ 8.5199995 ,  8.88000011,  8.55000019,  7.94999981,  8.60999966,
          8.5199995 ,  8.80000019,  8.13000011,  8.68999958,  8.72999954,
          8.47999954,  8.25      ,  8.40999985,  8.43999958,  8.38999939,
          8.35999966,  8.63999939,  8.51000023,  8.36999989,  8.69999981,
          8.52999973,  8.13999939,  8.36999989,  8.42000008,  8.55999947,
          8.72999954,  9.09000015,  8.18999958,  8.76000023,  8.53999996]]], dtype=float32)

Unfortunately when I called plt.contour on this, it said again "Input must be a 2D array".

Is there something I have missed?

Thanks,

Jamie

[toc] | [prev] | [next] | [standalone]

#76372

Jamie Mitchell wrote:

> I created the 2D array which read as:

That's not a 2D array.

When the amount of data you have is too big to clearly see what it
happening, replace it with something smaller. Instead of 30 items per
sub-array, try it with 5 items per sub-array. Instead of eight decimal
places, try it with single-digit integers. Anything to make it small enough
to see clearly.

When I do that with your data, instead of this:

> array([[[ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
>           2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
>           2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
>           1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
>           2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
>           2.16600013,  2.00999999,  1.86200011,  2.19800019, 
>           2.01200008]],
> 
>        [[ 8.5199995 ,  8.88000011,  8.55000019,  7.94999981,  8.60999966,
>           8.5199995 ,  8.80000019,  8.13000011,  8.68999958,  8.72999954,
>           8.47999954,  8.25      ,  8.40999985,  8.43999958,  8.38999939,
>           8.35999966,  8.63999939,  8.51000023,  8.36999989,  8.69999981,
>           8.52999973,  8.13999939,  8.36999989,  8.42000008,  8.55999947,
>           8.72999954,  9.09000015,  8.18999958,  8.76000023, 
>           8.53999996]]], dtype=float32)


I get this:


array([[[ 2,  2,  2,  1,  2]],
       [[ 8,  8,  8,  7,  8]]], dtype=float32)


which is much easier to work with. See the difference between that smaller
example, and my earlier explanation of the difference between a 1D and 2D
array?

One dimensional arrays are made from a single list of numbers: [...]
Two dimensional arrays are made from a list of lists: [ [...], [...] ]

*Three* dimensional arrays are made from a list of lists of lists: 
[ [ [...], [...] ] ]

*Four* dimensional arrays are made from a list of lists of lists of lists:
[ [ [ [...], [...] ] ] ]

and so on. You have a 3D array, with dimensions 2 x 1 x 30.

You can check the dimensions by storing the array into a variable like this:

py> a = numpy.array([[[ 2,  2,  2,  1,  2]], [[ 8,  8,  8,  7,  8]]])
py> a.shape
(2, 1, 5)



-- 
Steven

[toc] | [prev] | [next] | [standalone]

#76485

On Friday, August 15, 2014 4:13:26 PM UTC+1, Steven D'Aprano wrote:
> Jamie Mitchell wrote:
> 
> 
> 
> > I created the 2D array which read as:
> 
> 
> 
> That's not a 2D array.
> 
> 
> 
> When the amount of data you have is too big to clearly see what it
> 
> happening, replace it with something smaller. Instead of 30 items per
> 
> sub-array, try it with 5 items per sub-array. Instead of eight decimal
> 
> places, try it with single-digit integers. Anything to make it small enough
> 
> to see clearly.
> 
> 
> 
> When I do that with your data, instead of this:
> 
> 
> 
> > array([[[ 2.08800006,  2.29400015,  2.00400019,  1.88000011,  2.0480001 ,
> 
> >           2.16800022,  2.0480001 ,  1.88200009,  1.95800006,  2.00200009,
> 
> >           2.02800012,  1.81200004,  1.95000005,  1.96200013,  1.95200014,
> 
> >           1.99800014,  2.07000017,  1.88200009,  1.98400009,  2.13400006,
> 
> >           2.11400008,  1.89400005,  2.05000019,  2.01999998,  2.03400016,
> 
> >           2.16600013,  2.00999999,  1.86200011,  2.19800019, 
> 
> >           2.01200008]],
> 
> > 
> 
> >        [[ 8.5199995 ,  8.88000011,  8.55000019,  7.94999981,  8.60999966,
> 
> >           8.5199995 ,  8.80000019,  8.13000011,  8.68999958,  8.72999954,
> 
> >           8.47999954,  8.25      ,  8.40999985,  8.43999958,  8.38999939,
> 
> >           8.35999966,  8.63999939,  8.51000023,  8.36999989,  8.69999981,
> 
> >           8.52999973,  8.13999939,  8.36999989,  8.42000008,  8.55999947,
> 
> >           8.72999954,  9.09000015,  8.18999958,  8.76000023, 
> 
> >           8.53999996]]], dtype=float32)
> 
> 
> 
> 
> 
> I get this:
> 
> 
> 
> 
> 
> array([[[ 2,  2,  2,  1,  2]],
> 
>        [[ 8,  8,  8,  7,  8]]], dtype=float32)
> 
> 
> 
> 
> 
> which is much easier to work with. See the difference between that smaller
> 
> example, and my earlier explanation of the difference between a 1D and 2D
> 
> array?
> 
> 
> 
> One dimensional arrays are made from a single list of numbers: [...]
> 
> Two dimensional arrays are made from a list of lists: [ [...], [...] ]
> 
> 
> 
> *Three* dimensional arrays are made from a list of lists of lists: 
> 
> [ [ [...], [...] ] ]
> 
> 
> 
> *Four* dimensional arrays are made from a list of lists of lists of lists:
> 
> [ [ [ [...], [...] ] ] ]
> 
> 
> 
> and so on. You have a 3D array, with dimensions 2 x 1 x 30.
> 
> 
> 
> You can check the dimensions by storing the array into a variable like this:
> 
> 
> 
> py> a = numpy.array([[[ 2,  2,  2,  1,  2]], [[ 8,  8,  8,  7,  8]]])
> 
> py> a.shape
> 
> (2, 1, 5)
> 
> 
> 
> 
> 
> 
> 
> -- 
> 
> Steven

Thanks for your suggestions Steven. Unfortunately I still can't make the plot I'm looking for.

Do you mind if I go back to the start? Sorry I'm probably not explaining what I need very well.

So I have two 1D arrays:

1st array - ([8, 8.8,8.5,7.9,8.6 ...], dtype=float32)

It has a shape (150,)

2nd array - ([2, 2.2, 2.5, 2.3, ...],dtype=float32)

It has a shape (150,)

What I want to do is create a 2D array which merges the 1st and 2nd array so that I would have:

([[8, 8.8,8.5,7.9,8.6 ...],[2,2,2,2,5,2.3, ...]], dtype=float32) that would have a shape (150,150)

In this form I could then plot a 2D contour.

Thanks for your patience.

Jamie

[toc] | [prev] | [next] | [standalone]

#76487

I forgot to mention that when I try:

a=np.array([[hs_con_sw],[te_con_sw]])

I get a 3D shape for some reason - (2,1,150) which is not what I'm after.

Thanks,

Jamie

[toc] | [prev] | [next] | [standalone]

#76494

On Monday, August 18, 2014 10:25:15 PM UTC+5:30, Jamie Mitchell wrote:
> I forgot to mention that when I try:

> a=np.array([[hs_con_sw],[te_con_sw]])

> I get a 3D shape for some reason - (2,1,150) which is not what I'm after.

I guess you want

a=np.array([hs_con_sw,te_con_sw])

??

[toc] | [prev] | [next] | [standalone]

#76528

Jamie Mitchell wrote:

> I forgot to mention that when I try:
> 
> a=np.array([[hs_con_sw],[te_con_sw]])
> 
> I get a 3D shape for some reason - (2,1,150) which is not what I'm after.

No need to wrap the arrays hs_con_sw and te_con_sw in [] lists, since
they're already arrays.

a = np.array([hs_con_sw, te_con_sw])

ought to do what you want. That's a list [] of arrays, hence
two-dimensional.



-- 
Steven

[toc] | [prev] | [next] | [standalone]

#76543

You were right Christian I wanted a shape (2,150).

Thank you Rustom and Steven your suggestion has worked.

Unfortunately the data doesn't plot as I imagined.

What I would like is:

X-axis - hs_con_sw
Y-axis - te_con_sw
Z-axis - Frequency

What I would like is for the Z-axis to contour the frequency or amount of times that the X-axis data and Y-axis data meet at a particular point or bin.

Does anyone know what function or graph could best show this?

Thanks for all your help,

Jamie

[toc] | [prev] | [next] | [standalone]

#76563

On Tue, 19 Aug 2014 01:43:54 -0700 (PDT), Jamie Mitchell
<jamiemitchell1604@gmail.com> declaimed the following:

>You were right Christian I wanted a shape (2,150).
>
>Thank you Rustom and Steven your suggestion has worked.
>
>Unfortunately the data doesn't plot as I imagined.
>
>What I would like is:
>
>X-axis - hs_con_sw
>Y-axis - te_con_sw
>Z-axis - Frequency
>
>What I would like is for the Z-axis to contour the frequency or amount of times that the X-axis data and Y-axis data meet at a particular point or bin.
>
>Does anyone know what function or graph could best show this?
>
	Sounds like a histogram on a function of (x, y) -- you may need to play
with it to get tolerances suitable for binning, and calculate the Z by
scanning the input data.

	What are the domain ranges of X and Y? How many bins in X and Y?

initialize Z array [x-bins, y-bins] <- 0
for (x, y) in zip(x-data, y-data)
	convert x and y to bin indices
	increment Z[x-index, y-index]
plot Z (mapped to a color scale, dot size, whatever) on a grid of x-bins,
y-bins

-- 
	Wulfraed                 Dennis Lee Bieber         AF6VN
    wlfraed@ix.netcom.com    HTTP://wlfraed.home.netcom.com/

[toc] | [prev] | [next] | [standalone]

#76605

Jamie Mitchell <jamiemitchell1604@gmail.com> writes:

> You were right Christian I wanted a shape (2,150).
>
> Thank you Rustom and Steven your suggestion has worked.
>
> Unfortunately the data doesn't plot as I imagined.
>
> What I would like is:
>
> X-axis - hs_con_sw
> Y-axis - te_con_sw
> Z-axis - Frequency
>
> What I would like is for the Z-axis to contour the frequency or
> amount of times that the X-axis data and Y-axis data meet at a
> particular point or bin.
>
> Does anyone know what function or graph could best show this?

in my understanding, you have 3 arrays of data that describe 3D data
points, and you want to draw a 2D contour plot...

in this case you have to interpolate the z-values on a regular grid,
that's very easy if you already know what to do ;-)

here I assume that data is in a .csv file

% cat a.csv
0 ≤ x ≤ 10, 0 ≤ y ≤ 10, z = cos(sqrt((x-5)**2_(y-5)**2))
1.922065,5.827944,-0.998953
7.582322,0.559370,0.411861
5.001753,3.279957,-0.148694
...

of course my z's are different from yours, but this shouldn't be a
real problem --- and here it is my *tested* solution (tested on python
2.7, that is), please feel free to adapt to your needs

hth, ciao
                                                       g

% cat contour.py
from numpy import loadtxt, linspace
from matplotlib.mlab import griddata
import matplotlib.pyplot as pl

# open 'a.csv', specify the delimiter, specify how many header rows,
# slurp the data
temp_array = loadtxt(open('a.csv'),delimiter=',',skiprows=1)

# the shape of temp_array is (N,3), we want its transpose
temp_array = temp_array.transpose()

# now the shape is (3,N) and we can do "unpack and assignment:
x, y, z = temp_array

# now the tricky part, 

# 1: create two arrays with 101 (arbitrary number) equispaced values
# between 0 and 10 --- that is the ranges of data x and data y
xi = linspace(0,10,101)
yi = linspace(0,10,101)

# 2: create, by interpolation, the 2D array that contourf so eagerly
# awaited!
print griddata.__doc__
zi = griddata(x,y,z, xi,yi)

# eventually, lets plot the stuff...
# see http://matplotlib.org/examples/pylab_examples/griddata_demo.html
# for further details and ideas

pl.contour (xi,yi,zi,11,linewidths=1,colors='black')
pl.contourf(xi,yi,zi); pl.colorbar()
# optional
pl.gca().set_aspect('equal', 'box')
pl.show()
% python contour.py

[toc] | [prev] | [next] | [standalone]

#76651

On Tuesday, August 19, 2014 10:21:48 PM UTC+1, pec...@pascolo.net wrote:
> Jamie Mitchell <jamiemitchell1604@gmail.com> writes:
> 
> 
> 
> > You were right Christian I wanted a shape (2,150).
> 
> >
> 
> > Thank you Rustom and Steven your suggestion has worked.
> 
> >
> 
> > Unfortunately the data doesn't plot as I imagined.
> 
> >
> 
> > What I would like is:
> 
> >
> 
> > X-axis - hs_con_sw
> 
> > Y-axis - te_con_sw
> 
> > Z-axis - Frequency
> 
> >
> 
> > What I would like is for the Z-axis to contour the frequency or
> 
> > amount of times that the X-axis data and Y-axis data meet at a
> 
> > particular point or bin.
> 
> >
> 
> > Does anyone know what function or graph could best show this?
> 
> 
> 
> in my understanding, you have 3 arrays of data that describe 3D data
> 
> points, and you want to draw a 2D contour plot...
> 
> 
> 
> in this case you have to interpolate the z-values on a regular grid,
> 
> that's very easy if you already know what to do ;-)
> 
> 
> 
> here I assume that data is in a .csv file
> 
> 
> 
> % cat a.csv
> 
> 0 ≤ x ≤ 10, 0 ≤ y ≤ 10, z = cos(sqrt((x-5)**2_(y-5)**2))
> 
> 1.922065,5.827944,-0.998953
> 
> 7.582322,0.559370,0.411861
> 
> 5.001753,3.279957,-0.148694
> 
> ...
> 
> 
> 
> of course my z's are different from yours, but this shouldn't be a
> 
> real problem --- and here it is my *tested* solution (tested on python
> 
> 2.7, that is), please feel free to adapt to your needs
> 
> 
> 
> hth, ciao
> 
>                                                        g
> 
> 
> 
> % cat contour.py
> 
> from numpy import loadtxt, linspace
> 
> from matplotlib.mlab import griddata
> 
> import matplotlib.pyplot as pl
> 
> 
> 
> # open 'a.csv', specify the delimiter, specify how many header rows,
> 
> # slurp the data
> 
> temp_array = loadtxt(open('a.csv'),delimiter=',',skiprows=1)
> 
> 
> 
> # the shape of temp_array is (N,3), we want its transpose
> 
> temp_array = temp_array.transpose()
> 
> 
> 
> # now the shape is (3,N) and we can do "unpack and assignment:
> 
> x, y, z = temp_array
> 
> 
> 
> # now the tricky part, 
> 
> 
> 
> # 1: create two arrays with 101 (arbitrary number) equispaced values
> 
> # between 0 and 10 --- that is the ranges of data x and data y
> 
> xi = linspace(0,10,101)
> 
> yi = linspace(0,10,101)
> 
> 
> 
> # 2: create, by interpolation, the 2D array that contourf so eagerly
> 
> # awaited!
> 
> print griddata.__doc__
> 
> zi = griddata(x,y,z, xi,yi)
> 
> 
> 
> # eventually, lets plot the stuff...
> 
> # see http://matplotlib.org/examples/pylab_examples/griddata_demo.html
> 
> # for further details and ideas
> 
> 
> 
> pl.contour (xi,yi,zi,11,linewidths=1,colors='black')
> 
> pl.contourf(xi,yi,zi); pl.colorbar()
> 
> # optional
> 
> pl.gca().set_aspect('equal', 'box')
> 
> pl.show()
> 
> % python contour.py

This is great and works very well - thank you!!

[toc] | [prev] | [next] | [standalone]

#76676

On 20/08/2014 11:02, Jamie Mitchell wrote:
>
> This is great and works very well - thank you!!
>

I'm pleased to see that you have answers.  In return would you please 
access this list via 
https://mail.python.org/mailman/listinfo/python-list or read and action 
this https://wiki.python.org/moin/GoogleGroupsPython to prevent us 
seeing double line spacing and single line paragraphs, thanks.

-- 
My fellow Pythonistas, ask not what our language can do for you, ask
what you can do for our language.

Mark Lawrence

[toc] | [prev] | [next] | [standalone]

#76498

Am 18.08.14 18:51, schrieb Jamie Mitchell:
> On Friday, August 15, 2014 4:13:26 PM UTC+1, Steven D'Aprano wrote:
> So I have two 1D arrays:
>
> 1st array - ([8, 8.8,8.5,7.9,8.6 ...], dtype=float32)
>
> It has a shape (150,)
>
> 2nd array - ([2, 2.2, 2.5, 2.3, ...],dtype=float32)
>
> It has a shape (150,)
>
> What I want to do is create a 2D array which merges the 1st and 2nd array so that I would have:
>
> ([[8, 8.8,8.5,7.9,8.6 ...],[2,2,2,2,5,2.3, ...]], dtype=float32) that would have a shape (150,150)

There is a logical error in your description. This array would not have 
the shape (150,150), but (2,150).

> In this form I could then plot a 2D contour.

An image with 150x150 pixels has in total 150x150=22500 pixels. You've 
got only 300 datapoints. So it's very unclear what you want to do. A 
"contour plot" displays lines of equal height over a two-dimensional 
dataset, as in a topographic map. You don't seem to have a 
two-dimensional dataset at first place; maybe you already have *contour 
data*, and you just want to plot(x,y) ?

	Christian

[toc] | [prev] | [next] | [standalone]

#76468

Jamie Mitchell <jamiemitchell1604@gmail.com> writes:

> I created the 2D array which read as:

Maybe you could try numpy.reshape() on your 1D array?

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web