从 scipy interpolate/griddata 检索数据点

Question

I used Scipy's Griddata to fill in this gridded data using the points plotted (shown as empty). Is there a way to get the interpolated values (z) based on (x,y) coordinates? here's the code for the plot, and the x,y,z values are all Series.

    xi = np.linspace(min(lats),max(lats),360)
    yi = np.linspace(min(lons),max(lons),360)
    # grid the data.
    zi = griddata((lats, lons), nuits, (xi[None,:], yi[:,None]), method='cubic')
    # contour the gridded data.
    plt.contourf(xi,yi,zi,15,cmap=cMap)
    plt.colorbar()
    # plot data points.
    #plt.scatter(lats,lons,c=nuits,marker='o',s=26,cmap=cMap)
    plt.scatter(lats,lons,facecolors='none', edgecolors='k',s=26)
    plt.show()

Answer 1

This would work:

xi_coords = {value: index for index, value in enumerate(xi)}
yi_coords = {value: index for index, value in enumerate(yi)}
xic = <your coordinate>
yic = <your coordinate>
zi[xi_coords[xic], yi_coords[yic]]

Answer 2

You can get your interpolated zi coordinates from your (xi,yi) coordinates by:

# (xi,yi) coords to get the interpolated zi coords
# where len(xic) = len(yic)  
xic = <your coordinate>
yic = <your coordinate>

# sort these coordinates in an increasing order
s_xic = np.sort(xic)
s_yic = np.sort(yic)

# indices belonging to xic, yic, that would sort the array
ind_s_xic = np.argsort(xic)
ind_s_yic = np.argsort(yic)

 
dict_xic = dict(zip(ind_s_xic, np.array(range(len(xic))))
dict_yic = dict(zip(ind_s_yic, np.array(range(len(yic))))


xi,yi = np.meshgrid(s_xic, s_yic)

# zi_grid has dimensions ( len(yic), len(xic) )
zi_grid = griddata((lats, lons), nuits, (xi, yi), method='cubic')

# zic is the interpolated z-coordinate data with an arrangement order,
# corresponding to the x and y-coordinate data in xic and yic
zic =  [ zi_grid[dict_yic[i], dict_xic[i]] for i in range(len(xic)) ]

Visit How does one use numpy's Meshgrid function with a random interval rather than a equally spaced one? to understand about how meshgrid works.

Meshgrid can be created from your unevenly spaced (xi,yi) coordinates, following which, griddata can use your meshgrid to interpolate the z-coords from the interpolated surface created based on points = (lats, lons), values = nuits.