在matplotlib中自定義x軸

Question

在下圖中，x軸上的每個單位表示一個10分鐘的間隔。我想自定義X軸的標籤，以便它顯示小時，即每6個單位（60分鐘）顯示一個標籤。我是matplotlib的新手。有人能幫助我嗎？謝謝〜

這是上圖的代碼。

x = arange(0, size_x, dx) 
y = arange(0, size_y, dy) 
X,Y = meshgrid(x, y) 
Z = foo(x,y) 
pcolor(X, Y, Z, cmap=cm.Reds) 
colorbar() 
axis([0,size_x-1,0,size_y-1]) 
show() 

Answer 1

有多種方法可以做到這一點。

讓我們先用一個例子情節：

import matplotlib.pyplot as plt 
import matplotlib as mpl 
import numpy as np 

# Generate some data... 
x, y = np.mgrid[:141, :101] 
z = np.cos(np.hypot(x, y)) 

# Plot the figure... 
plt.pcolormesh(x, y, z, cmap=mpl.cm.Reds) 

plt.show() 

最簡單的方法做你想要會是這樣的：

import matplotlib.pyplot as plt 
import matplotlib as mpl 
import numpy as np 

# Generate some data... 
x, y = np.mgrid[:141, :101] 
z = np.cos(np.hypot(x, y)) 

# Plot the figure... 
plt.pcolormesh(x, y, z, cmap=mpl.cm.Reds) 

# Set the ticks and labels... 
ticks = np.arange(x.min(), x.max(), 6) 
labels = range(ticks.size) 
plt.xticks(ticks, labels) 
plt.xlabel('Hours') 

plt.show() 

另一種方式涉及到子標籤唱matplotlib的定位器和代碼。

出於您的目的，上述示例沒有問題。

製作新的定位器和代碼的好處在於，該軸將自動縮放到您指定的「dx」單位的合理間隔中。如果您將其用作較大應用程序的一部分，則可能是值得的。對於一個情節，它比它的價值更麻煩。

如果你真的想要走這條路，雖然，你會做這樣的事情：

import matplotlib.pyplot as plt 
import matplotlib as mpl 
import numpy as np 

def main(): 
    # Generate some data... 
    x, y = np.mgrid[:141, :101] 
    z = np.cos(np.hypot(x, y)) 

    # Plot the figure... 
    fig, ax = plt.subplots() 
    ax.pcolormesh(x, y, z, cmap=mpl.cm.Reds) 
    ax.set_xlabel('Hours') 

    ax.xaxis.set_major_locator(ScaledLocator(dx=6)) 
    ax.xaxis.set_major_formatter(ScaledFormatter(dx=6)) 

    plt.show() 

class ScaledLocator(mpl.ticker.MaxNLocator): 
    """ 
    Locates regular intervals along an axis scaled by *dx* and shifted by 
    *x0*. For example, this would locate minutes on an axis plotted in seconds 
    if dx=60. This differs from MultipleLocator in that an approriate interval 
    of dx units will be chosen similar to the default MaxNLocator. 
    """ 
    def __init__(self, dx=1.0, x0=0.0): 
     self.dx = dx 
     self.x0 = x0 
     mpl.ticker.MaxNLocator.__init__(self, nbins=9, steps=[1, 2, 5, 10]) 

    def rescale(self, x): 
     return x/self.dx + self.x0 
    def inv_rescale(self, x): 
     return (x - self.x0) * self.dx 

    def __call__(self): 
     vmin, vmax = self.axis.get_view_interval() 
     vmin, vmax = self.rescale(vmin), self.rescale(vmax) 
     vmin, vmax = mpl.transforms.nonsingular(vmin, vmax, expander = 0.05) 
     locs = self.bin_boundaries(vmin, vmax) 
     locs = self.inv_rescale(locs) 
     prune = self._prune 
     if prune=='lower': 
      locs = locs[1:] 
     elif prune=='upper': 
      locs = locs[:-1] 
     elif prune=='both': 
      locs = locs[1:-1] 
     return self.raise_if_exceeds(locs) 

class ScaledFormatter(mpl.ticker.OldScalarFormatter): 
    """Formats tick labels scaled by *dx* and shifted by *x0*.""" 
    def __init__(self, dx=1.0, x0=0.0, **kwargs): 
     self.dx, self.x0 = dx, x0 

    def rescale(self, x): 
     return x/self.dx + self.x0 

    def __call__(self, x, pos=None): 
     xmin, xmax = self.axis.get_view_interval() 
     xmin, xmax = self.rescale(xmin), self.rescale(xmax) 
     d = abs(xmax - xmin) 
     x = self.rescale(x) 
     s = self.pprint_val(x, d) 
     return s 

if __name__ == '__main__': 
    main()