Problem on the results given by the code

[youssefhakam]

if running the program, the code gives good results but when adding streamlit the results are faulty.

import numpy as np
import streamlit as st
st.title("""App localization""")

import numpy as np
import matplotlib.pyplot as plt
class HMM(object):

    def __init__(self,Transition_Matrix,initial_state):
        self.Transition_Matrix  = Transition_Matrix
        self.initial_state = initial_state 
        pass

    def filtering_estimation(self, Observ):
        new_state_No_Normalise = np.dot(Observ,np.dot(self.Transition_Matrix,self.initial_state ))
        new_state = new_state_No_Normalise/np.linalg.norm(new_state_No_Normalise)
        self.initial_state = new_state
        return new_state

    def Obseravation_matrix(self,error,no_discrepancies):
        sensor_list = []
        lenght = len (self.initial_state)
        Obsevation_Matrix = np.zeros((lenght,lenght))
        for d in no_discrepancies:
            proba = ((1-error)**(4-d))*(error**d)
            sensor_list.append(proba)
        np.fill_diagonal(Obsevation_Matrix,sensor_list)
        return Obsevation_Matrix
    
    def state_actuel(self,filtering_estimation):
        state_actuel = np.argwhere(filtering_estimation == np.amax(filtering_estimation)) + 1 
        return state_actuel.flatten().tolist()
        #print("state actuel possible ", state_actuel.flatten().tolist())

#definition des variable

transition_matrix = np.array([
    [0.16, 0.84, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
    [0.28, 0.16, 0.28, 0.28, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
    [0, 0.84, 0.16, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
    [0, 0, 0, 0.16, 0, 0, 0, 0, 0, 0, 0.84, 0, 0, 0, 0],
    [0, 0, 0, 0, 0.16, 0.42, 0, 0, 0, 0.42, 0, 0, 0, 0, 0],
    [0, 0, 0, 0, 0.42, 0.16, 0, 0, 0, 0.42, 0, 0, 0, 0, 0],
    [0 , 0, 0, 0, 0, 0, 0.16, 0.84, 0, 0, 0, 0, 0, 0, 0],
    [0, 0, 0, 0, 0, 0, 0.28, 0.16, 0.28, 0, 0, 0, 0, 0, 0.28],
    [0, 0, 0, 0, 0, 0.21, 0, 0.21, 0.16, 0.21, 0, 0, 0, 0.21, 0],
    [0, 0, 0, 0, 0.21, 0, 0, 0, 0.21, 0.16, 0.21, 0, 0.21, 0, 0],
    [0, 0.21, 0, 0.21, 0, 0, 0, 0, 0, 0.21, 0.16, 0.21, 0, 0, 0],
    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.42, 0.16, 0.42, 0, 0],
    [0, 0, 0, 0, 0, 0, 0, 0, 0, 0.42, 0, 0.42, 0.16, 0, 0],
    [0, 0, 0, 0, 0, 0, 0, 0, 0.84, 0, 0, 0, 0, 0.16, 0],
    [0, 0, 0, 0, 0, 0, 0, 0.84, 0, 0, 0, 0, 0, 0, 0.16]

]) 

initial_state = np.array([1,0,0,0,0,0,0,0,0,0,0,0,0,0,0])

Model = HMM(transition_matrix,initial_state)

observation_matrix_SWE = Model.Obseravation_matrix(0.2,[0,2,2,2,3,1,2,2,4,4,4,1,1,0,0])
observation_matrix_NWE = Model.Obseravation_matrix(0.2,[2,2,0,0,1,1,0,2,4,4,4,3,3,2,2])
observation_matrix_W = Model.Obseravation_matrix(0.2,[2,0,2,2,1,3,2,2,2,2,2,1,3,3,2,2])
observation_matrix_E = Model.Obseravation_matrix(0.2,[2,2,2,2,3,1,2,0,2,2,2,3,1,2,2])
observation_matrix_SW = Model.Obseravation_matrix(0.2,[1,1,3,3,2,4,3,3,3,3,3,0,2,1,1])
observation_matrix_SE = Model.Obseravation_matrix(0.2,[1,3,3,3,4,2,3,1,3,3,3,2,0,1,1])
observation_matrix_NW = Model.Obseravation_matrix(0.2,[3,1,1,1,0,2,1,3,3,3,3,2,4,3,3])
observation_matrix_NE = Model.Obseravation_matrix(0.2,[2,3,1,1,2,0,1,1,3,3,3,4,2,3,3])
observation_matrix_ND = Model.Obseravation_matrix(0.2,[4,2,4,4,3,3,4,2,0,0,0,3,3,4,4])


#Dectionnaire 

Dictionnaire_Of_Obseravation = {
    "SWE" : observation_matrix_SWE,
    "NWE" : observation_matrix_NWE,
    "W" : observation_matrix_W,
    "E" : observation_matrix_E,
    "SW" : observation_matrix_SW,
    "SE" : observation_matrix_SE,
    "NW" : observation_matrix_NW,
    "ND" :  observation_matrix_ND,
    "NE" : observation_matrix_NE
    
} 
options = st.selectbox(
     'What are your favorite colors',
     ['SWE', 'NWE', 'W', 'E', 'SW', 'SE', 'NW', 'NE', 'ND']
     )

st.write('You selected:', options)
state = Model.filtering_estimation(Dictionnaire_Of_Obseravation[options])
#print("state actuel possible ", Model.state_actuel(state))

#fig, ax = plt.subplots()
#s = state.tolist()
    #print(state)
#ax.bar(range(1,len(s)+1), s)
#plt.xticks(range(1, 16))

#st.pyplot(fig)
st.bar_chart(state)
st.write('distribution initial:', state )

[mliu]

@youssefhakam can you describe what results you are seeing that is faulty? I am seeing consistent results on state and Model.state_actual(state)

[youssefhakam]

thanks for responding.
When I don’t use Streamlit, the program gives me the correct state, but when I use Streamlit, it gives me the incorrect state.
example this is an image of state without using streamlit

and this image of state using streamlit

The correct result is the first

[Goyo]

With which selection?

This is what I get running your code, selection SWE:

An this without using streamlit:

Looks the same to me.

[youssefhakam]

Thanks for responding.
Ok try the sequence of observation W and NWE

[Goyo]

W, using Streamlit

W, without Streamlit

NWE, using Streamlit

NWE, without Streamlit