app.py

import numpy as np
import matplotlib.pylab as plt
import ruptures as rpt
import streamlit as st
from ruptures.metrics import precision_recall
from ruptures.metrics import hausdorff
from ruptures.metrics import randindex


st.title("Change Point Detection")
# Generating Signal

def pw_constant_input(n,dim,n_bkps,sigma):
    """Piecewise constant (pw_constant)"""
    # n, dim  # number of samples, dimension
    # n_bkps, sigma # number of change points, noise standard deviation
    signal, bkps = rpt.pw_constant(n, dim, n_bkps, noise_std=sigma)
    rpt.display(signal, bkps)
    return signal,bkps

def pw_linear_input(n,dim,n_bkps,sigma):
    """Piecewise Linear"""
    # creation of data
    # n, dim = 500, 3  # number of samples, dimension of the covariates
    # n_bkps, sigma = 3, 5  # number of change points, noise standart deviation
    signal, bkps = rpt.pw_linear(n, dim, n_bkps, noise_std=sigma)
    rpt.display(signal, bkps)
    return signal,bkps

def pw_normal_input(n,dim,n_bkps,sigma):
    """Piecewise 2D Gaussian process (pw_normal)#"""
    # creation of data
    #n = 500  # number of samples
    #n_bkps = 3  # number of change points
    signal, bkps = rpt.pw_normal(n, n_bkps)
    rpt.display(signal, bkps)
    return signal,bkps

def pw_wavy_input(n,dim,n_bkps,sigma):
    # creation of data
    #n, dim = 500, 3  # number of samples, dimension
    #n_bkps, sigma = 3, 5  # number of change points, noise standart deviation
    signal, bkps = rpt.pw_wavy(n, n_bkps, noise_std=sigma)
    rpt.display(signal, bkps)
    return signal,bkps

input_list = ['piecewiseConstant','piecewiseLinear','piecewiseNormal','piecewiseSinusoidal']
generate_signal = st.selectbox(label = "Choose an input signal", options = input_list)

n,dim,n_bkps,sigma = st.columns(4)
with n:
    n= st.number_input('No of Samples',min_value=100,step=1)
with dim:
    dim = st.number_input('No of dimesions',min_value=1,max_value = 5,step=1)
with n_bkps:
    n_bkps = st.number_input('No of breakpoints',min_value=2,step=1)
with sigma:
    sigma = st.number_input('Variance',min_value=0,max_value=4,step=1)

if generate_signal == 'piecewiseConstant':
     signal,bkps = pw_constant_input(n,dim,n_bkps,sigma)
elif generate_signal== 'piecewiseLinear':
    signal,bkps = pw_linear_input(n,dim,n_bkps,sigma)
elif generate_signal == 'piecewiseNormal':
    signal,bkps = pw_normal_input(n,dim,n_bkps,sigma)
else:
    signal,bkps= pw_wavy_input(n,dim,n_bkps,sigma)

fig, axarr = rpt.display(signal,bkps)
st.pyplot(fig)

def dynp_method(signal,bkps,n_bkps):
    # change point detection
    model = "l1"  # "l2", "rbf"
    algo = rpt.Dynp(model=model, min_size=3, jump=5).fit(signal)
    my_bkps = algo.predict(n_bkps)
    # show results
    fig,axarr = rpt.show.display(signal, bkps, my_bkps, figsize=(10, 2))
    #plt.show()
    st.pyplot(fig)
    return my_bkps

def pelt_method(signal,bkps,n_bkps):
    # change point detection
    model = "l1"  # "l2", "rbf"
    algo = rpt.Pelt(model=model, min_size=n_bkps, jump=5).fit(signal)
    my_bkps = algo.predict(pen=n_bkps)

    # show results
    fig, ax_arr = rpt.display(signal, bkps, my_bkps, figsize=(10, 2))
    st.pyplot(fig)
    return my_bkps
    

def bin_seg_method(signal,bkps,n_bkps):
    # change point detection
    model = "l2"  # "l1", "rbf", "linear", "normal", "ar",...
    algo = rpt.Binseg(model=model).fit(signal)
    my_bkps = algo.predict(n_bkps)

    # show results
    fg,axxarr = rpt.show.display(signal, bkps, my_bkps, figsize=(10, 2))
    st.pyplot(fig)
    return my_bkps

def bot_up_seg(signal,bkps,n_bkps):
    # change point detection
    model = "l2"  # "l1", "rbf", "linear", "normal", "ar",...
    algo = rpt.Binseg(model=model).fit(signal)
    my_bkps = algo.predict(n_bkps)

    # show results
    fig,axxar = rpt.show.display(signal, bkps, my_bkps, figsize=(10, 2))
    st.pyplot(fig)
    return my_bkps

def win_sli_seg(signal,bkps,n_bkps):
    # change point detection
    model = "l2"  # "l1", "rbf", "linear", "normal", "ar"
    algo = rpt.Window(width=40, model=model).fit(signal)
    my_bkps = algo.predict(n_bkps)

    # show results
    fig,axxar= rpt.show.display(signal, bkps, my_bkps, figsize=(10, 2))
    st.pyplot(fig)
    return my_bkps
    

searchmethod_list = ['Dynamic Programming','Pelt','Binary Segmentation','Bottom-up Segmentation','Window sliding segmentation']
detection_model = st.selectbox(label = "Choose a Detection Method",options = searchmethod_list)

if detection_model== 'Dynamic Programming':
    bkps1 = dynp_method(signal,bkps,n_bkps)
    
elif detection_model=='Pelt':
    bkps1 = pelt_method(signal,bkps,n_bkps)
elif detection_model=='Binary Segmentation':
    bkps1 = bin_seg_method(signal,bkps,n_bkps)
elif detection_model=='Bottom-up Segmentation':
    bkps1 = bot_up_seg(signal,bkps,n_bkps)
else:
    bkps1 = win_sli_seg(signal,bkps,n_bkps)

p, r = precision_recall(bkps, bkps1)
st.header('Precision and Recall')
st.write(p, r)

st.header('Hausdorff metric')
st.write(hausdorff(bkps, bkps1))
    
st.header('Rand index')


st.write(randindex(bkps, bkps1))



st.write("""
For a detailed description please look through our Documentation 
""")

url = 'https://huggingface.co/spaces/ThirdEyeData/ChangePointDetection/blob/main/README.md'

st.markdown(f'''
<a href={url}><button style="background-color: #668F45;">Documentation</button></a>
''',
unsafe_allow_html=True)