streamlit-app/app.py at main · thom1100/streamlit-app · GitHub

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
import streamlit as st
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go

@st.cache_data
def load_data(train_path, test_path):
    train = pd.read_parquet(train_path)
    test = pd.read_parquet(test_path)
    return train, test

def main():
    st.set_page_config(page_title="Bike Counters Dashboard", layout="wide")
    st.title("🚴‍♂️ Bike Counters Dashboard")
    st.markdown("Une exploration interactive des données de comptage de vélos.")

    train_path = "data/train.parquet"
    test_path = "data/final_test.parquet"
    train, test = load_data(train_path, test_path)

    st.header("📋 Aperçu des données")
    st.write("**Train set** :", train.shape, " | **Test set** :", test.shape)
    st.dataframe(train.head())

    st.sidebar.header("🔍 Filtres")
    if "site_name" in train.columns:
        selected_site = st.sidebar.selectbox("Choisir un compteur :", sorted(train["site_name"].unique()))
        filtered_df = train[train["site_name"] == selected_site]
    else:
        selected_site = None
        filtered_df = train

    if "date" in train.columns:
        train["date"] = pd.to_datetime(train["date"])
        min_date, max_date = train["date"].min().date(), train["date"].max().date()
        selected_date = st.sidebar.slider("Choisir une plage de dates :", min_date, max_date, (min_date, max_date))
        filtered_df = filtered_df[(filtered_df["date"] >= pd.to_datetime(selected_date[0])) & (filtered_df["date"] <= pd.to_datetime(selected_date[1]))]

    st.header("📈 Visualisations interactives")

    if "bike_log_count" in filtered_df.columns:
        fig_hist = px.histogram(
            filtered_df,
            x="bike_log_count",
            nbins=40,
            color="site_name" if "site_name" in filtered_df.columns else None,
            title="Distribution du nombre de vélos (log count)"
        )
        st.plotly_chart(fig_hist, use_container_width=True)

    if "date" in filtered_df.columns and "bike_log_count" in filtered_df.columns:
        fig_time = px.line(
            filtered_df,
            x="date",
            y="bike_log_count",
            color="site_name" if selected_site is None else None,
            title=f"Évolution du trafic vélo {f'pour {selected_site}' if selected_site else ''}"
        )
        st.plotly_chart(fig_time, use_container_width=True)

    numeric_cols = filtered_df.select_dtypes(include="number").columns
    if len(numeric_cols) > 1:
        corr = filtered_df[numeric_cols].corr()
        fig_corr = px.imshow(corr, text_auto=True, color_continuous_scale="Viridis", title="Matrice de corrélation")
        st.plotly_chart(fig_corr, use_container_width=True)

    st.header("📊 Statistiques descriptives")
    st.write(filtered_df.describe())

    st.header("🧪 Données de test")
    st.dataframe(test.head())

if __name__ == "__main__":
    main()