File size: 3,433 Bytes
4b88786
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import importlib


from st_on_hover_tabs import on_hover_tabs

import streamlit as st

import streamlit_presentation
import streamlit_presentation.analyse
importlib.reload(streamlit_presentation.analyse)
from streamlit_presentation.analyse import repartition_par_categorie
from streamlit_presentation.analyse import repartition_longueur_categorie


import streamlit_presentation.preprocessing
importlib.reload(streamlit_presentation.preprocessing)
from streamlit_presentation.preprocessing import detection_langage_et_traduction

import streamlit_presentation.modele
importlib.reload(streamlit_presentation.modele)
from streamlit_presentation.modele import presentation_modele
from sklearn.metrics import f1_score

plt.rcParams['font.size'] = 12
plt.rcParams['axes.labelsize'] = 10
plt.rcParams['axes.titlesize'] = 12
plt.rcParams['xtick.labelsize'] = 8
plt.rcParams['ytick.labelsize'] = 8
plt.rcParams['legend.fontsize'] = 8
plt.rcParams['lines.linewidth'] = 1

#on charge les donnees utilisees
data = pd.read_csv( 'data.csv')
extract_data = pd.read_csv( 'data_tr_extract.csv')
sum_data = pd.read_csv( 'data_sum_extract.csv')
test_data = pd.read_pickle( 'data_test.pkl')

from keras.models import load_model
import tensorflow as tf
from tensorflow.keras import backend as K
import ast


def f1_weighted(true, pred):  

    # Classes
    classes = K.arange(0, 27) 
    true = K.one_hot(K.cast(true, 'int32'), 27)
    
    # Calcule les TP, FP, FN pour chaque classe
    tp = K.dot(K.transpose(true), K.round(pred))
    fp = K.dot(K.transpose(1-true), K.round(pred))
    fn = K.dot(K.transpose(true), 1-K.round(pred))

    # Calcule le score F1 pour chaque classe
    p = tp / (tp + fp + K.epsilon())
    r = tp / (tp + fn + K.epsilon())
    f1 = 2*p*r / (p+r+K.epsilon())

    
    weighted_f1 = K.sum(f1 * K.sum(true, axis=0) / K.sum(true))
    return weighted_f1

model = load_model("final_model_kfold.h5", custom_objects={'f1_weighted': f1_weighted})





from sklearn.preprocessing import LabelEncoder
encoder =  LabelEncoder()
print(test_data.columns)
y_test = encoder.fit_transform(test_data["prdtypecode"])
class_labels = encoder.classes_
label_size = 27



####### Page principale
st.set_page_config(layout="wide")
st.markdown('<style>' + open('./style.css').read() + '</style>', unsafe_allow_html=True)

st.title("Mon Application")

with st.sidebar:
    tabs = on_hover_tabs(tabName=['Introduction', "Analyse", "Preprocessing", "Modèle", "Pistes exploratoires"], 
                         iconName=['apps', 'bar_chart', "sync", "memory", "topic"], default_choice=0)

st.markdown("""
<style>
    .rounded-border-parent {
        border-radius: 15px !important;
        border: 1px solid blue !important;
        background-color: lightgray !important;
    }
</style>
""", unsafe_allow_html=True)


if tabs == "Introduction":
    st.write("# Introduction")
    st.write("Ici")

elif tabs == "Analyse":
    st.write("# Analyse")
    
    st.dataframe(data.head(30))
    st.write("")
    
    repartition_par_categorie(st, data)
    repartition_longueur_categorie(st, data)

elif tabs == "Preprocessing":
    detection_langage_et_traduction(st, extract_data, sum_data)

elif tabs == "Modèle":
    presentation_modele(st, test_data, model,class_labels,y_test)

elif tabs == "Pistes exploratoires":
    st.write("# Pistes exploratoires")
    st.write("Ici")