pythonseabornkdeplot

How to find last minima in density curve

i can't find easy way to get minima related to a density curve. With this kind of data :

import matplotlib.pyplot as plt
from scipy.misc import electrocardiogram
from scipy.signal import find_peaks
import seaborn as sns
    test = [7, 7, 7, 7, 8, 7, 7, 7, 7, 7, 7, 8, 8, 8, 7, 7, 7, 8, 7, 7, 8, 7, 7, 8, 7, 7, 8, 7, 8, 7, 7, 8, 7, 8, 8, 8, 8, 7, 7, 7, 7, 7, 7, 8, 7, 4, 4, 7, 8, 7]
    sns.kdeplot(test)

enter image description here

i would like to get the last inflection point related to the last "peak", in this situation it would be a minima around x=8. I tried function like get_peaks but this is not doing the job because i need to get all the values after this last inflection point and in some situation, like this one :

test2 = [8, 7, 7, 8, 7, 7, 7, 7, 7, 7, 7, 7, 7, 8, 7, 7, 7, 7, 7, 6, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 8, 7, 7, 7, 7, 7, 8, 12]

get_peaks didn't give me the last number.

My idea would be to get the x position of the last minima and after being able to get all the number higher than the minima.

Solution

  • It is better to place your latest find_peaks() implementation. Here I am trying to get it using that function.

    import matplotlib.pyplot as plt
from scipy.signal import find_peaks
import seaborn as sns

test = [7, 7, 7, 7, 8, 7, 7, 7, 7, 7, 7, 8, 8, 8, 7, 7, 7, 8, 7, 7, 8, 7, 7, 8, 7, 7, 8, 7, 8, 7, 7, 8, 7, 8, 8, 8, 8, 7, 7, 7, 7, 7, 7, 8, 7, 4, 4, 7, 8, 7]
ax = sns.kdeplot(test)
x = ax.lines[0].get_xdata() # Get the x data of the distribution
y = ax.lines[0].get_ydata() # Get the y data of the distribution
xy = [[x[i], y[i]] for i in range(len(x))]

# find peak
peak_coord = [xy[i] for i in find_peaks(y)[0]]
sorted_peak = sorted(peak_coord, key=lambda x: x[1])

# sort peak based on its `y` coord
sorted_peak.reverse() # [[7.01539631380498, 0.7002187588539351], [7.921442255354041, 0.3685891140800782], [3.984828854140883, 0.043144074377544736]]

# second peak
second_peak = sorted_peak[1] # [7.921442255354041, 0.3685891140800782]

plt.show()

    As you can see, I can get the second peak is in coord [7.921442255354041, 0.3685891140800782]