如何将模型中带有颜色的数据绘制到matplotlib底图上？ 为什么没有显示数据？

Question

I am trying to get my data output from my model plotted on Basemap. I can see the map, but not the points. What is the issue? I am using the latest version of scikit-learn and basemap for python 2.7. I have the following code:

dataframe = pd.read_csv('powerplants.csv') # CSV file that contains latitude column, longitude column, and id for the powerplant
colormap = np.arange(500)
labels = modeloutput.labels_ # assume this is output of my model

fig = plt.figure()
ax = fig.add_subplot(111, axisbg='w', frame_on=True)
fig.set_size_inches(18.5, 10.5)
map_points = [...] # assume this is a list populated with coordinates from the csv
# Plot the clusters on the map
# m is a basemap object
m.scatter(
         [geom.x for geom in map_points],
         [geom.y for geom in map_points],
         20, marker='o', lw=.25,
         c = colormap(labels.astype(float)),
         alpha =0.9, antialiased=True,
         zorder=3)
m.fillcontinents(color='#555555')
plt.show()

Answer 1

This example is modified from the official example . Key parts for using colormap in scatter plot are:

1. x.data and y.data are transformed coordinates.

2. c = np.random.randint(1, 500, size=len(lats)) are values mapping to a color in the colormap corresponding to each point.

Some parts which might be not necessary for you are:

 import urllib, os from netCDF4 import Dataset import numpy as np filename, _ = urllib.urlretrieve('http://coastwatch.pfeg.noaa.gov/erddap/tabledap/apdrcArgoAll.nc?longitude,latitude,time&longitude>=0&longitude<=360&latitude>=-90&latitude<=90&time>=2010-01-01&time<=2010-01-08&distinct()') dset = Dataset(filename) lats = dset.variables['latitude'][:] lons = dset.variables['longitude'][:] dset.close() os.remove(filename) c = np.random.randint(1, 500, size=len(lats)) x, y = m(lons,lats) 

This part is used for generating an data sample. You might want to replace it with your real data.

 from mpl_toolkits.axes_grid1 import make_axes_locatable divider = make_axes_locatable(ax) fig.colorbar(pc, cax=divider.append_axes("right", size="5%", pad=0.05)) 

This is a direct application of bogatron's answer to make the size of colorbar matches the plot. It's your choice of keeping it or not.

import urllib, os

import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import make_axes_locatable
from mpl_toolkits.basemap import Basemap
from netCDF4 import Dataset
import numpy as np

# data downloaded from the form at
# http://coastwatch.pfeg.noaa.gov/erddap/tabledap/apdrcArgoAll.html
filename, _ = urllib.urlretrieve('http://coastwatch.pfeg.noaa.gov/erddap/tabledap/apdrcArgoAll.nc?longitude,latitude,time&longitude>=0&longitude<=360&latitude>=-90&latitude<=90&time>=2010-01-01&time<=2010-01-08&distinct()')
dset = Dataset(filename)
lats = dset.variables['latitude'][:]
lons = dset.variables['longitude'][:]
dset.close()
os.remove(filename)
c = np.random.randint(1, 500, size=len(lats))

# draw map with markers for float locations
fig = plt.figure()
ax = fig.add_subplot(111, axisbg='w', frame_on=True)
fig.set_size_inches(18.5, 10.5)
m = Basemap(lon_0=180, ax=ax)
x, y = m(lons,lats)
pc = m.scatter(x.data, y.data, 20, marker='o', c=c, lw=.25, alpha =0.9, antialiased=True, zorder=3, cmap='summer')
m.fillcontinents(color='#555555')
divider = make_axes_locatable(ax)
fig.colorbar(pc, cax=divider.append_axes("right", size="5%", pad=0.05))
plt.show()