app.py

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

with open('/Users/zsapey/Documents/Code/libe-datalab/Sciences-POC/output_favarel_et_al.txt', 'r') as f : 
    out_dict = json.loads(f.read())
    
@st.cache_resource
def initialize():

    articles = pd.read_csv('/Users/zsapey/Documents/Code/libe-datalab/Sciences-POC/extract_sciences_po.csv')

    with open('/Users/zsapey/Documents/Code/libe-datalab/Sciences-POC/output_favarel_et_al.txt', 'r') as f : 
        out_dict = json.loads(f.read())
    
    df = pd.DataFrame.from_dict(out_dict)

    articles = pd.merge(df, articles, on='item_id', how='left')

    count_principale = df.groupby('categorie_principale').item_id.count()
    count_secondaire = df.groupby('categorie_secondaire').item_id.count()
    display_principale = count_principale.reset_index()
    display_principale.columns = ['Catégorie', 'Nombre d\'articles']
    display_secondaire = count_secondaire.reset_index()
    display_secondaire.columns = ['Catégorie', 'Nombre d\'articles']

    template ="ggplot2"

    fig = go.Figure()
        
    fig.update_layout(template=template, 
                      )

    fig.add_trace(go.Scatterpolar(
        r=display_principale['Nombre d\'articles'],
        theta=display_principale['Catégorie'],
        fill='toself',
        name='Catégorie Principale',
        marker = {'color' : 'red'},

    ))
    fig.add_trace(go.Scatterpolar(
        r=display_secondaire['Nombre d\'articles'],
        theta=display_secondaire['Catégorie'],
        fill='toself',
        name='Catégorie Secondaire',
        marker = {'color' : 'blue'},
        opacity=0.3,
    ))

    fig.update_layout(
    polar=dict(
        radialaxis=dict(
        visible=True,
        range=[0, max(max(display_principale['Nombre d\'articles']), max(display_secondaire['Nombre d\'articles']))]
        )),
    showlegend=True
    )
    fig.update_layout(legend=dict(
    yanchor="top",
    y=0.0001,
    xanchor="left",
    x=0.395
))
    return fig, display_principale, articles


def display_article(article):

    url = article['url']

    colImage, colText = st.columns(2)
# try :   
    with colImage :
        st.image(article["image_url"]) # image URL
    with colText:
        if 'subhead' in article.index and article['subhead']!='nan':
            st.subheader(f":red[{article['subhead']}] [{article['titre'].rstrip('Libération').rstrip('-')[:-2]}]({url})") # Title 
        else : 
            # st.toast(article.index)
            titre_cleaned = article['titre'].removesuffix('Libération').rstrip('-').strip()
            st.subheader(f"[{titre_cleaned}]({url})") # Title 
        st.write(f"{article['description']}") # Header
        formatted_date = article["date_published"]
        if article.premium:
            st.markdown(
        f"""
        <span style='color:grey'>{formatted_date+" "} </span>  <span style='color:#eeb54e'> abonnés</span>
    """,
    unsafe_allow_html=True
    )
        else :
            st.markdown(
        f"""
        <span style='color:grey'>{formatted_date+" "} </span>
    """,
    unsafe_allow_html=True
    )
    # except :
    #     st.toast(f'Error displaying article {article.item_id}')
    #     return 
        

fig, display_principale, articles = initialize()
# col1, col2, col3 = st.columns([0.5, 0.2, 0.3])

st.subheader('Répartition des articles par catégorie')
# with col1:
col1, col2 = st.columns([0.6, 0.4], vertical_alignment='center')

with col1:
    st.plotly_chart(fig)

with col2:
    st.dataframe(display_principale.set_index('Catégorie').sort_values(by='Nombre d\'articles', ascending=False))

st.subheader('Exemples d\'articles')
tabs = st.tabs(display_principale['Catégorie'].values.tolist())

with tabs[0]:
    cat = display_principale['Catégorie'][0]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)


with tabs[1]:
    cat = display_principale['Catégorie'][1]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)


with tabs[2]:
    cat = display_principale['Catégorie'][2]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)


with tabs[3]:
    cat = display_principale['Catégorie'][3]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)


with tabs[4]:
    cat = display_principale['Catégorie'][4]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)


with tabs[5]:
    cat = display_principale['Catégorie'][5]
    for i, article in articles.loc[articles.categorie_principale==cat].iterrows():
        display_article(article)