Feat/ajout graphs pgp

eu_fact_force/dash-app/pgpxd4g_graphs.py

@@ -0,0 +1,131 @@
+import pandas as pd
+import plotly.graph_objects as go
+
+DATA_PATH = "data/PGP x D4G- Exported Vaccine Data.xlsx" # Correct to the actual Excel file
+
+# ── Load all sheets ────────────────────────────────────────────────
+xl = pd.ExcelFile(DATA_PATH)
+
+df_trend = pd.read_excel(xl, "Trendline")
+df_trend.columns = ["date", "posts"]
+df_trend["date"] = pd.to_datetime(df_trend["date"])
+
+df_lang = pd.read_excel(xl, "Language")
+df_lang.columns = ["language", "posts", "share"]
+df_lang = df_lang.dropna(subset=["language", "posts"])
+df_lang = df_lang[df_lang["posts"].apply(lambda x: str(x).isdigit() or isinstance(x, (int, float)))]
+df_lang["posts"] = pd.to_numeric(df_lang["posts"], errors="coerce")
+df_lang = df_lang.dropna(subset=["posts"]).head(8)
+
+df_source = pd.read_excel(xl, "Source")
+df_source.columns = ["source", "mentions"]
+df_source["mentions"] = pd.to_numeric(df_source["mentions"], errors="coerce").fillna(0)
+df_source = df_source.dropna(subset=["source", "mentions"])
+
+df_themes_raw = pd.read_excel(xl, "Crosstab- themes across all vac")
+
+# Corrected processing for df_th
+df_th = df_themes_raw.drop(columns=['Topics \\ Themes']).T.reset_index()
+df_th.columns = ['theme', 'mentions']
+df_th['mentions'] = pd.to_numeric(df_th['mentions'], errors='coerce')
+df_th = df_th.dropna(subset=["mentions"])
+df_th = df_th.sort_values("mentions", ascending=False)
+
+PALETTE = ["#1E5AA8", "#2F6FB6", "#5C8FD6", "#E6EEF8"]
+BG = "#FFFFFF"
+PAPER = "#FFFFFF"
+FONT_COLOR = "#333333"
+GRID_COLOR = "#E0E0E0"
+
+
+def base_layout(title):
+    return dict(
+        title=dict(text=title, font=dict(size=16, color=FONT_COLOR), x=0.02),
+        paper_bgcolor=PAPER,
+        plot_bgcolor=BG,
+        font=dict(family="Inter, sans-serif", color=FONT_COLOR),
+        margin=dict(l=50, r=30, t=60, b=50),
+    )
+
+
+# ── 1. Daily post volume trendline ─────────────────────────────
+fig_trend = go.Figure()
+fig_trend.add_trace(go.Scatter(
+    x=df_trend["date"], y=df_trend["posts"],
+    mode="lines+markers",
+    line=dict(color=PALETTE[0], width=2),
+    marker=dict(size=5, color=PALETTE[0]),
+    fill="tozeroy",
+    fillcolor="rgba(30,90,168,0.15)",
+    name="Posts",
+    hovertemplate="%{x|%b %d}<br><b>%{y:,}</b> posts<extra></extra>",
+))
+fig_trend.update_layout(
+    **base_layout("Daily Post Volume — March–April 2026"),
+    xaxis=dict(showgrid=False, tickformat="%b %d", tickcolor=GRID_COLOR),
+    yaxis=dict(showgrid=True, gridcolor=GRID_COLOR, tickformat=","),
+)
+
+# ── 2. Language donut ───────────────────────────────────
+fig_lang = go.Figure(go.Pie(
+    labels=df_lang["language"],
+    values=df_lang["posts"],
+    hole=0.55,
+    marker=dict(colors=PALETTE),
+    textinfo="label+percent",
+    hovertemplate="<b>%{label}</b><br>%{value:,} posts (%{percent})<extra></extra>",
+))
+fig_lang.update_layout(
+    **base_layout("Post Distribution by Language"),
+    showlegend=False,
+)
+
+# Create a color list for df_source that matches its length
+source_colors = []
+for i in range(len(df_source)):
+    source_colors.append(PALETTE[i % len(PALETTE)])
+
+# ── 3. Source donut chart ───────────────────────────────────
+fig_source = go.Figure(go.Pie(
+    labels=df_source["source"].tolist(),
+    values=df_source["mentions"].tolist(),
+    hole=0.55,
+    marker=dict(colors=source_colors), # Use the explicit color list
+    textinfo="label+percent", # Reverted to include percentage
+    hovertemplate="<b>%{label}</b><br>%{value:,} mentions (%{percent})<extra></extra>", # Reverted to include percentage
+))
+fig_source.update_layout(
+    **base_layout("Mentions by Platform"),
+    showlegend=False,
+)
+
+# ── 4. Anti-vaccine themes treemap ─────────────────────
+# Define a colorscale using the PALETTE in reverse order to map higher mentions to darker blues
+gradient_colors = [[0, PALETTE[3]], [0.33, PALETTE[2]], [0.66, PALETTE[1]], [1, PALETTE[0]]]
+fig_themes = go.Figure(go.Treemap(
+    labels=df_th["theme"],
+    parents=[""] * len(df_th),
+    values=df_th["mentions"],
+    # Assign the 'mentions' column to marker.colors for the gradient effect
+    marker=dict(colors=df_th["mentions"], colorscale=gradient_colors, showscale=False),
+    hovertemplate="<b>%{label}</b><br>%{value:,} mentions (%{percent parent})<extra></extra>",
+    textinfo="label+percent parent"
+))
+fig_themes.update_layout(**base_layout("Anti-vaccine Themes — All Vaccines"))
+fig_themes.update_layout(margin=dict(l=10, r=10, t=60, b=10))
+
+# ── Preview all figures ────────────────────────────────
+if __name__ == "__main__":
+    for name, fig in [
+        ("Trendline", fig_trend),
+        ("Language", fig_lang),
+        ("Source", fig_source),
+        ("Themes Treemap", fig_themes)
+    ]:
+        print(f"Attempting to display: {name}")
+        try:
+            fig.show()
+            print(f"Opened: {name}")
+        except Exception as e:
+            print(f"Error opening {name}: {e}")
+            print(f"Skipping {name} due to error.")

visualisation_trendline.py

@@ -0,0 +1,33 @@
+# trendline data
+file_path = "PGP x D4G- Exported Vaccine Data.xlsx" # excel source to be modified 
+
+trendline = pd.read_excel(file_path, sheet_name="Trendline")
+
+# colors creation
+primary = "#1E5AA8"
+secondary = "#2F6FB6"
+background = "#E6EEF8"
+
+# graph creation
+import plotly.graph_objects as go
+
+fig = go.Figure()
+
+fig.add_trace(go.Scatter(
+    x=trendline["Publication Date (GMT+01:00) London"],
+    y=trendline["Posts"],
+    name="Posts count evolution",
+    line=dict(color=primary, width=2),
+    marker=dict(size=6, color=secondary),
+    hovertemplate="Date: %{x}<br>Posts: %{y}<extra></extra>"
+))
+
+fig.update_layout(
+    title="Posts count evolution in the last month",
+    plot_bgcolor="white",
+    paper_bgcolor=background,
+    xaxis_title="Date",
+    yaxis_title="Number of posts",
+)
+
+fig.show()