將圖形中的線條標籤保留在縮放時可見

Question

我正在使用matplotlibs plt.axvline在圖形上繪製垂直線條。我想將垂直線上的標籤直接放到圖上，而不是放在圖例中。

我希望我的圖表有看起來像這樣的垂直線標籤：

如果我放大某個部分，我想垂直線標籤移動。因此，使他們仍然可以在屏幕上看到。就像這樣：

誰能幫助我？這將非常感謝！

到目前爲止我的代碼低於：

from astropy.io import fits 
from astropy.utils.data import download_file 
from astropy.table import Table 
import matplotlib.pyplot as plt 
import numpy as np 

class Spectrum: 
    # Let's download the data to plot  
    def __init__(self, url): 
     self.url = url 
     self.hdu_list = fits.open(download_file(url, cache=True), memmap=False) 

    # Now lets plot the data 
    def plot_spectra(self): 
     x = np.array(Table(self.hdu_list[1].data).columns[1]) 
     x = 10 ** x 
     y = np.array(Table(self.hdu_list[1].data).columns[0]) 
     plt.plot(x, y, 'k', lw=1) 

    # Now lets plot the vertical lines, AND THIS IS WHERE I WANT TO ADD LABELS. 
    def plot_spectral_types(self): 
     my_type = input("Please enter the spectral type to plot (o, b, a, or f): ") 
     if my_type is 'o': 
      my_type = o_type 
     elif my_type is 'b': 
      my_type = b_type 
     elif my_type is 'a': 
      my_type = a_type 
     elif my_type is 'f': 
      my_type = f_type 

     element, wavelength = zip(*my_type) 
# Each vertical line's x value is a wavelength. 
# I want the vertical line's label to be the corresponding element. 
     for i in wavelength: 
      plt.axvline(linewidth=0.25, color='r', x=i) 

o_type = [ 
('NIII', 4097), 
('SiIV', 4089), 
('H', 4340.5), 
('HeI', 4471), 
('HeII', 4541), 
('NIII', 4632), 
('NIII', 4640), 
('CIII', 4650), 
('HeII', 4686) 
] 

b_type = [ 
('SiIV', 4089), 
('H', 4101.7), 
('HeI', 4121), 
('SiII', 4128), 
('SiII', 4131), 
('H', 4340.5), 
('HeI', 4471), 
('CIII', 4540), 
('HeII', 4541), 
('CIII', 4650), 
('H', 4861.33) 
] 

a_type = [ 
('CaII (K)', 3933.70), 
('CaII', 3968.50), 
('H', 3970.10), 
('H', 4101.70), 
('HeI', 4121.00), 
('SiII', 4128.00), 
('SiII', 4131.00), 
('FeI', 4299.00), 
('FeI', 4303.00), 
('TiII', 4303.00), 
('H', 4340.50), 
('MgII', 4481.00), 
('H', 4861.30), 
('H', 6562.70) 
] 

f_type = [ 
('CaII', 3933.70), 
('CaII', 3968.50), 
('H', 3970.10), 
('H', 4101.70), 
('HeI', 4121.00), 
('SiII', 4128.00), 
('SiII', 4131.00), 
('CaI', 4227.00), 
('FeI', 4299.00), 
('FeI', 4303.00), 
('H', 4340.50), 
('CH', 4314.00), 
('MgII', 4481.00), 
('H', 4861.30), 
('H', 6562.70) 
] 

Answer 1

文字標籤在data-coordinates默認情況下，這可能是你所遇到的問題。您可以將y值轉換爲圖形座標，它在放大時相對於當前座標軸，並將x座標保留在數據座標中。更獨立的（見MCVE）示例：

from numpy.random import uniform 
from math import sin, pi 
import matplotlib.pyplot as plt 
import matplotlib.transforms as transforms 



fig, ax = plt.subplots() 

transDA = transforms.blended_transform_factory(
    ax.transData, ax.transAxes) # from the transforms tutorial 

spectrum = uniform(0,1, 1000) + map(lambda x: sin(2*pi*x/800), range(1000)) #dummy data 

ax.plot(range(4000,5000,1), spectrum) 

o_type = [ 
('NIII', 4097), 
('SiIV', 4089), 
('H', 4340.5), 
('HeI', 4471), 
('HeII', 4541), 
('NIII', 4632), 
('NIII', 4640), 
('CIII', 4650), 
('HeII', 4686) 
] 

for wavelength in o_type: 
     print(wavelength[0],wavelength[1]) 
     plt.axvline(linewidth=0.25, color='r', x=wavelength[1]) 
     plt.text(wavelength[1], # x-value from data 
       uniform(0,1), # wiggle the labels 2so they don't overlap 
       wavelength[0], # string label 
       transform = transDA, 
       color='red', 
       family='serif') # the III looked better serif. 

原始情節。請注意，前兩個標籤分別位於圖的上半部分，可能是圖片的四分之一處：
放大，這些標籤在放大的x軸上正常移動，但仍位於圖的一半左右，一個四分之一的數字：