How to draw a vertical line at the mode of the seaborn distplot

Question

I just learned how to draw a density plot with the seaborn Python module:

import numpy as np
import torch
from matplotlib import pyplot as plt
from matplotlib.pyplot import (plot, savefig, xlim, figure,
                              ylim, legend, boxplot, setp,
                              axes, xlabel, ylabel, xticks,
                              axvline)
import seaborn as sns

layer1_G1_G2 = [-0.05567627772688866,
 -0.06829605251550674,
 -0.0721447765827179,
 -0.05942181497812271,
 -0.061410266906023026,
 -0.062010858207941055,
 -0.05238522216677666,
 -0.057129692286252975,
 -0.06323938071727753,
 -0.07018601894378662,
 -0.05972284823656082,
 -0.06124034896492958,
 -0.06971242278814316,
 -0.06730005890130997]

def make_density(layer_list,color, layer_num):

    layer_list_tensor = torch.tensor(layer_list)
    
    # Plot formatting
    plt.title('Density Plot of Median Stn. MC-Losses at Layer ' + layer_num)
    plt.xlabel('MC-Loss')
    plt.ylabel('Density')
    plt.xlim(-0.2,0.05)
    plt.ylim(0, 85)
    min_ylim, max_ylim = plt.ylim()
    
    # Draw the density plot
    sns.distplot(layer_list, hist = False, kde = True,
                 kde_kws = {'linewidth': 2}, color=color)

# plot the density plot
# the resulting density plot is shown below
>>> make_density(layer1_G1_G2, 'green','1')

How can I draw a vertical line at the mode of this density curve on this distplot?

Thank you,

For calculating the mode of the list, check this https://stackoverflow.com/questions/10797819/finding-the-mode-of-a-list — Miguel Trejo, Aug 29 '20 at 16:51
For the vertical line, this may be useful https://stackoverflow.com/questions/52334938/seaborn-how-to-add-vertical-lines-to-a-distribution-plot-sns-distplot — Miguel Trejo, Aug 29 '20 at 16:51

score 1 · Answer 1 · answered Aug 29 '20 at 17:02

I found the solution:

def make_density(layer_list,color, layer_num):

    
    # Plot formatting
    plt.title('Density Plot of Median Stn. MC-Losses at Layer ' + layer_num)
    plt.xlabel('MC-Loss')
    plt.ylabel('Density')
    plt.xlim(-0.2,0.05)
    plt.ylim(0, 85)
    min_ylim, max_ylim = plt.ylim()
    
    
    
    # Draw the density plot
    sns.distplot(layer_list, hist = False, kde = True,
                 kde_kws = {'linewidth': 2}, color=color)
    
    dens_list = sns.distplot(layer1_G1_G2, hist = False, kde = True,
             kde_kws = {'linewidth': 2}, color='green').get_lines()[0].get_data()[1].tolist()
                    
    max_dens_index = dens_list.index(max(dens_list))
    
    mode = sns.distplot(layer1_G1_G2, hist = False, kde = True,
             kde_kws = {'linewidth': 2}, color='green').get_lines()[0].get_data()[0].tolist()[max_dens_index]
  
    plt.axvline(mode, color='orange', linestyle='dashed', linewidth=1.5)

    plt.text(mode * 0.87, 80, 'mode: {:.2f}'.format(mode))

>>> make_density(layer1_G1_G2, 'green','1')

Note that by calling `sns.distplot` 3 times, the same plot is drawn 3 times one over the other. This can be avoided by storing the returned `ax` into a variable. — JohanC, Aug 29 '20 at 17:39

score 1 · Accepted Answer · answered Aug 29 '20 at 17:32

You can extract the x and y values of the generated curve and find the mode as the highest y-value.

from matplotlib import pyplot as plt
import seaborn as sns

layer1_G1_G2 = [-0.05567627772688866, -0.06829605251550674, -0.0721447765827179, -0.05942181497812271, -0.061410266906023026, -0.062010858207941055, -0.05238522216677666, -0.057129692286252975, -0.06323938071727753, -0.07018601894378662, -0.05972284823656082, -0.06124034896492958, -0.06971242278814316, -0.06730005890130997]

def make_density(layer_list, color, layer_num):
    # Draw the density plot
    ax = sns.distplot(layer_list, hist=False, kde=True, kde_kws={'linewidth': 2}, color=color)
    x = ax.lines[0].get_xdata()
    y = ax.lines[0].get_ydata()
    mode_idx = y.argmax()
    ax.vlines(x[mode_idx], 0, y[mode_idx], color='crimson', ls=':')

    # Plot formatting
    ax.set_title('Density Plot of Median Stn. MC-Losses at Layer ' + layer_num)
    ax.set_xlabel('MC-Loss')
    ax.set_ylabel('Density')
    ax.autoscale(axis='x', tight=True)
    ax.set_ylim(ymin=0)

make_density(layer1_G1_G2, 'green', '1')
plt.show()

How to draw a vertical line at the mode of the seaborn distplot

2 Answers2