How to plot numbers from an array as annotation using matplotlib?

Question

I am trying to produce the map on basemap using vales extracted from meteorological data. Sample code is:-

y=[2.56422, 3.77284,3.52623,3.51468,3.02199]
z=[0.15, 0.3, 0.45, 0.6, 0.75]
n=[58,651,393,203,123]

fig, ax = plt.subplots()
ax.scatter(z, y)

for i, txt in enumerate(n):
    ax.annotate(txt, (z[i],y[i]))

The data I am using is a numpy array. I dont know how to loop through each array to plot the kind of map similar to above. I would like to plot only values (ie. no countour or contourf).

Initially I was trying to plot float values using pylab.plot function. However, it retured with error

ValueError: third arg must be a format string

Then I tried to convert this numpy array to string and then plot with this command:-

temperature = np.array2string(data, precision=2)

and the print statement looks like a modified string:-

print temperature
[[ 19.69  21.09  21.57  21.45  20.59  20.53  20.93  20.63  20.64  21.26
   21.29  20.63  20.98  21.01  20.84  20.81  20.55  20.33  20.52  20.23
   19.84]
 [ 20.77  21.35  20.81  20.64  20.9   20.78  20.79  23.57  20.11  21.07
   21.06  21.33  21.48  21.18  21.4   21.09  20.5   20.31  20.12  19.8
   19.97]
 [ 21.51  21.23  20.55  20.08  20.05  20.78  21.17  24.77  21.17  20.95
   21.43  21.47  21.46  21.77  21.69  21.13  20.47  20.04  20.08  20.37
   20.14]
 [ 21.29  21.1   20.63  20.32  20.22  20.37  24.4   23.82  22.23  21.03
   22.11  22.62  22.71  22.37  21.73  21.35  21.03  20.67  20.58  20.89
   20.93]
 [ 21.24  21.04  20.68  20.56  20.76  20.91  24.26  23.75  23.28  21.26
   21.48  22.    21.94  21.78  21.36  21.14  20.96  20.92  21.1   21.19
   21.31]
 [ 20.83  20.88  20.6   20.87  21.01  21.91  22.33  22.21  21.74  20.66
   20.76  20.73  21.04  21.09  20.83  20.7   20.72  20.71  21.23  21.04
   20.73]
 [ 20.32  20.41  20.19  20.05  20.68  22.17  21.82  20.67  19.85  19.02
   18.91  19.6   20.15  20.64  20.64  20.09  19.81  19.76  19.9   19.94
   19.46]
 [ 19.68  20.37  20.56  20.68  20.93  21.28  21.24  20.33  20.7   20.
   18.72  18.94  19.56  19.57  19.83  19.74  19.17  18.53  18.1   18.72
   19.12]
 [ 18.88  19.71  20.77  20.81  20.32  21.58  20.96  21.33  21.2   20.17
   19.95  22.05  19.72  19.85  19.3   18.75  18.69  18.44  17.57  17.2
   18.22]
 [ 19.11  19.19  20.13  20.78  21.25  21.98  21.15  20.96  20.66  20.14
   20.51  21.92  20.36  20.27  19.    18.22  17.81  17.58  17.16  16.67
   17.46]
 [ 18.5   19.28  19.57  20.01  21.16  21.01  21.06  20.93  20.62  19.89
   20.3   20.7   19.7   19.76  18.24  17.    16.36  16.63  17.62  17.32
   17.38]
 [ 17.6   18.33  20.27  19.97  20.63  20.51  21.09  21.39  20.81  19.55
   20.    18.3   17.32  18.24  17.57  17.15  16.42  15.76  16.14  16.45
   21.95]
 [ 17.04  17.55  18.16  18.32  21.23  20.5   20.41  19.82  20.7   20.55
   20.41  18.47  18.05  17.63  17.11  15.6   16.02  15.46  14.29  13.88
   23.04]]

Finally, I get this error when I tried to plot the above value on a map with this line

pylab.plot(x, y, temperature)

    'Unrecognized character %c in format string' % c)
ValueError: Unrecognized character [ in format string

Problem seems to be with nparray to string conversion.

Any help to solve this issue is appreciated.

Could you please explain the structure of your data? Do you have a regular matrix with temperature values for every pixel? Could you provide some file as example to make some test? — Alejandro, Apr 15 '16 at 19:16
data --> http://www.mediafire.com/download/0epjjdm8auit611/mslp.txt x --> http://www.mediafire.com/download/1dn6p8nw96h2mmd/xlong.txt y --> http://www.mediafire.com/download/31suzsz6j7u2bgz/xlat.txt. The first link is not a temperature value but array length is same. — sundar_ima, Apr 15 '16 at 19:19
What is the structure of the data? Why I get arrays of shape (178, 343)? — Alejandro, Apr 15 '16 at 19:31
The working example I have posted earlier is here http://stackoverflow.com/questions/35633421/how-to-remove-omit-smaller-contour-lines-using-matplotlib/. The code is two lines below 'Full working example code for removing lines:-' — sundar_ima, Apr 15 '16 at 19:35
You can see my improved question with a very simple example. — sundar_ima, Apr 15 '16 at 20:16
What is not working in your example with `ax.annotate`? What's the problem with that? That seems to work. If the problem is with looping over a 2d array: just loop over `n.ravel()` instead of `n`. — Andras Deak -- Слава Україні, Apr 17 '16 at 23:35
Exactly what I wanted. Post with some code so that I can accept your answer. — sundar_ima, Apr 18 '16 at 08:30
@sundar_ima turns out there is an even better option than looping over `n.ravel()` and indexing, see my answer. — Andras Deak -- Слава Україні, Apr 18 '16 at 09:43

Andras Deak -- Слава Україні · Accepted Answer · 2022-01-26T09:57:14.170

Your original solution with ax.annotate is perfectly fine for your more general solution. The only thing to change is that in case of 2d arrays, you need to flatten them before looping over them using np.ravel() (which is also a method of the ndarray class).

However, in your specific case you can spare explicit indexing and the use of ravel() by broadcasting the three arrays you need to plot:

import numpy as np
import matplotlib.pyplot as plt

# generate some dummy data
rng = np.random.default_rng()
z, y = np.mgrid[:3, :3]
n = rng.integers(low=50, high=500, size=z.shape)

fig, ax = plt.subplots()
ax.scatter(z, y)

for zz, yy, txt in np.broadcast(z, y, n):
    ax.annotate(txt, (zz, yy))

Note that the result of np.broadcast is the same as if we'd used zip(z.ravel(), y.ravel(), n.ravel()).

Thanks for the answer. Works perfectly. – sundar_ima Apr 18 '16 at 13:50 — sundar_ima, Apr 18 '16 at 13:50

How to plot numbers from an array as annotation using matplotlib?

1 Answers1