Merge branch 'main' of https://github.com/Kirscher/ResultAthle

utils/viz.py

@@ -1,34 +1,61 @@
 import plotly.graph_objects as go
 import numpy as np
 import pandas as pd
-import utils.stat as stat
 
 from IPython.display import display, Markdown
 
-def graphePerso(prenom, nom, data, titre):
-    densite=stat.dens(data['duration'], bins = stat.idealBins(len(data['duration'])))
-    fcubic=stat.lissage(densite, sep = True,beginend = (data['duration'].min(),data['duration'].max()))
-    z=data.index[data.Athlète==(nom.upper()+" "+prenom.capitalize())]
-    temps = data.loc[z[0],'duration']
-    x= np.linspace(data['duration'].min(),data['duration'].max(), int(1e5))
+
+def graphePerso(prenom, nom, data, titre, nb_bins=10_000):
+    times_in_seconds = pd.to_datetime(data.h_duration, format='%H:%M:%S').dt.time.apply(
+        lambda time: 3600*time.hour+60*time.minute+time.second)
+    bins = np.linspace(times_in_seconds.iloc[0], times_in_seconds.iloc[-1], nb_bins)
+    cdf = [(times_in_seconds <= bin_).sum()/len(times_in_seconds) for bin_ in bins]
+
+    z = data.index[data['Athlète'] == (nom.upper() + " " + prenom.capitalize())]
+    temps = data.loc[z[0], 'duration']
+    cdf_proportion = (times_in_seconds <= temps).sum() / len(times_in_seconds)
 
     fig = go.Figure()
 
     fig.add_trace(go.Scatter(
-        x=x, y=fcubic(x),
+        x=bins, y=cdf,
         fill='tozeroy',
-        mode='lines', 
+        mode='lines',
         line_color='blue',
-        name='densité lissée & interpolée'
+        name='Distribution des temps'
     ))
 
     fig.add_trace(go.Scatter(
-        x=[temps, temps], y=[0, max(fcubic(x))],
+        x=[temps, temps], y=[0, max(cdf)],
         mode='lines',
         line_color='red',
         name=nom.upper()+" "+prenom.capitalize()
     ))
 
+    fig.add_shape(
+        type="line",
+        x0=bins[0],
+        y0=cdf_proportion,
+        x1=temps,
+        y1=cdf_proportion,
+        line=dict(
+            color="black",
+            width=1,
+            dash="dashdot",
+        ),
+    )
+
+    fig.add_annotation(
+        x=temps,
+        y=cdf_proportion,
+        text=f"<b>{100*cdf_proportion:.2f}%</b>",
+        font=dict(size=14),
+        showarrow=True,
+        arrowhead=7,
+        ax=0,
+        ay=-30,
+    )
+
     fig.update_layout(
         title=titre,
         xaxis_title="Durée pour franchir la ligne d'arrivée",
@@ -41,9 +68,49 @@ def graphePerso(prenom, nom, data, titre):
 
     fig.show()
 
+
 def display_header(header):
     display(Markdown(f"**Compétition:** {header['nom']}"))
     display(Markdown(f"**Lieu:** {header['lieu']}"))
     display(Markdown(f"**Date:** {header['date']}"))
     display(Markdown(f"**Dept:** {header['dept']}"))
-    display(Markdown(f"**Label:** {header['label']}"))
+    display(Markdown(f"**Label:** {header['label']}"))
+
+
+def display_podium(data):
+    data.h_duration = pd.to_datetime(data.h_duration, format='%H:%M:%S').dt.time
+    podium = data.sort_values(by='h_duration').iloc[[2, 0, 1]]
+
+    fig = go.Figure()
+
+    fig.add_trace(go.Bar(
+        y=[1, 3, 2],
+        x=podium['Athlète'],
+        orientation='v',
+        marker=dict(
+            color=['#614e1a', '#a57c00', '#c7d1da'],  # Couleur du podium (bronze, gold, silver)
+            line=dict(color='black', width=1)  # Couleur de la bordure
+        )
+    ))
+
+    for i, annotation_text in enumerate(['🥉', '🥇', '🥈']):
+        fig.add_annotation(
+            x=podium.iloc[i]['Athlète'],
+            y=[1.2, 3.2, 2.2][i],
+            text=annotation_text,
+            showarrow=False,
+            font=dict(size=70)
+        )
+
+    fig.update_layout(
+        title="Podium",
+        yaxis_title="Durée",
+        yaxis=dict(
+            tickvals=[1, 2, 3],
+            ticktext=podium['h_duration'].sort_values(ascending=False)
+        ),
+        xaxis=dict(
+            tickfont=dict(size=16)
+        )
+    )
+    fig.show()