Upload 3 files

app.py

@@ -0,0 +1,267 @@
+import joblib
+import numpy as np
+import gradio as gr
+
+# Lataa malli
+model = joblib.load("workload_model.joblib")
+
+CLASS_NAMES = {0: "Low", 1: "Medium", 2: "High"}
+EMOJI = {"Low": "🟢", "Medium": "🟠", "High": "🔴"}
+
+# Esimerkkipäivät dropdownille
+EXAMPLE_PRESETS = {
+    "Balanced day": [3, 2.5, 4, 2, 45, 9, 16],
+    "Overloaded day": [9, 7.5, 15, 0, 10, 9, 18],
+    "Focused day": [2, 1.5, 2, 3, 60, 8, 15],
+}
+
+
+def predict_workload(
+    meetings_count,
+    total_meeting_hours,
+    context_switches,
+    deep_work_blocks,
+    break_minutes,
+    day_start_hour,
+    day_end_hour,
+):
+    X = np.array([[
+        meetings_count,
+        total_meeting_hours,
+        context_switches,
+        deep_work_blocks,
+        break_minutes,
+        day_start_hour,
+        day_end_hour,
+    ]])
+
+    probs = model.predict_proba(X)[0]
+    pred_class = int(np.argmax(probs))
+    pred_label = CLASS_NAMES[pred_class]
+    confidence = float(probs[pred_class])
+
+    probs_dict = {name: float(probs[i]) for i, name in CLASS_NAMES.items()}
+
+    headline = (
+        f"<div style='text-align:center; font-size:1.8rem; margin:1rem 0;'>"
+        f"{EMOJI[pred_label]} <strong>{pred_label} Workload</strong>"
+        f"</div>"
+        f"<div style='text-align:center; color:#555; font-size:1.1rem;'>"
+        f"Confidence: <strong>{confidence * 100:.1f}%</strong>"
+        f"</div>"
+    )
+
+    # --- practical tips ---
+    tips = []
+    day_length = day_end_hour - day_start_hour
+
+    if meetings_count >= 8 or total_meeting_hours >= 6:
+        tips.append(
+            "Consider blocking <strong>meeting-free focus time</strong> (e.g. 2–3h) and moving "
+            "non-essential meetings to another day."
+        )
+    if context_switches >= 12:
+        tips.append(
+            "Try to <strong>batch similar tasks or meetings</strong> together to reduce context switching."
+        )
+    if deep_work_blocks == 0:
+        tips.append(
+            "Add at least <strong>one deep work block (60–90 minutes)</strong> for focused work without notifications."
+        )
+    if break_minutes < 30 and day_length >= 8:
+        tips.append(
+            "Increase <strong>break time</strong> slightly – even short 5–10 minute breaks every few hours reduce fatigue."
+        )
+    if day_length > 9:
+        tips.append(
+            "Your workday is long – consider <strong>moving low-priority tasks</strong> to another day or finishing earlier."
+        )
+
+    if not tips:
+        tips.append(
+            "Your setup looks balanced! To stay sustainable, consider scheduling regular deep work and micro-breaks."
+        )
+
+    tips_html = "<ul style='padding-left:1.2rem; line-height:1.6;'>"
+    for tip in tips:
+        tips_html += f"<li>{tip}</li>"
+    tips_html += "</ul>"
+
+    explanation = (
+        "<h3 style='margin-top:1.5rem;'>💡 Personalized Suggestions</h3>"
+        + tips_html +
+        "<h3 style='margin-top:1.5rem;'>🧠 How the Model Works</h3>"
+        "<p style='color:#555;'>This AI estimates your perceived workload using:</p>"
+        "<ul style='padding-left:1.2rem; color:#555;'>"
+        "<li>Number and duration of meetings</li>"
+        "<li>Frequency of task/meeting switches (context switches)</li>"
+        "<li>Presence of uninterrupted deep work blocks</li>"
+        "<li>Total break time during the day</li>"
+        "<li>Overall workday length</li>"
+        "</ul>"
+    )
+
+    return headline, probs_dict, explanation
+
+
+def load_example(example_name):
+    """Täyttää sliderit valitun esimerkin arvoilla."""
+    if example_name in EXAMPLE_PRESETS:
+        return EXAMPLE_PRESETS[example_name]
+    return [4, 3, 6, 1, 30, 9, 17]
+
+
+# Teema
+theme = gr.themes.Soft(
+    primary_hue=gr.themes.colors.orange,
+    secondary_hue=gr.themes.colors.rose,
+    neutral_hue=gr.themes.colors.slate,
+    radius_size=gr.themes.sizes.radius_md,
+).set(
+    button_primary_background_fill="*primary_500",
+    button_primary_background_fill_hover="*primary_600",
+    block_title_text_weight="600",
+)
+
+# CSS
+css = """
+#app-container {
+    max-width: 1200px;
+    margin: 0 auto;
+    padding: 0 1.5rem;
+}
+#app-header {
+    text-align: center;
+    margin-bottom: 2rem;
+}
+#app-header h1 {
+    font-weight: 700;
+    font-size: 2.2rem;
+    color: #222;
+    margin-bottom: 0.4rem;
+}
+#app-header p {
+    font-size: 1.1rem;
+    color: #666;
+    max-width: 650px;
+    margin: 0 auto;
+    line-height: 1.5;
+}
+.gr-button-primary {
+    font-weight: 600;
+    padding: 0.6rem 1.4rem;
+    font-size: 1.05rem;
+}
+@media (max-width: 768px) {
+    #app-header h1 {
+        font-size: 1.8rem;
+    }
+    #app-header p {
+        font-size: 1rem;
+    }
+    .gr-button-primary {
+        width: 100%;
+        padding: 0.7rem;
+    }
+}
+"""
+
+with gr.Blocks(theme=theme, css=css, title="🗓️ Workload Estimator") as demo:
+    with gr.Column(elem_id="app-container"):
+        gr.Markdown(
+            """
+            <div id="app-header">
+                <h1>🗓️ Calendar Workload Estimator</h1>
+                <p>Estimate your daily cognitive load based on meetings, focus time, and recovery.</p>
+            </div>
+            """,
+            elem_id="header"
+        )
+
+        with gr.Tab("📊 Estimate Your Day"):
+            with gr.Row(equal_height=False):
+
+                # VASEN SARake: dropdown + sliderit + nappi
+                with gr.Column(scale=2):
+                    example_dropdown = gr.Dropdown(
+                        label="💡 Load example schedule",
+                        choices=list(EXAMPLE_PRESETS.keys()),
+                        value=None,
+                        interactive=True,
+                    )
+
+                    gr.Markdown("### 🗓️ Workday Structure")
+                    meetings_count = gr.Slider(0, 12, value=4, step=1, label="Meetings (count)")
+                    total_meeting_hours = gr.Slider(0, 9, value=3, step=0.5, label="Total meeting hours")
+                    context_switches = gr.Slider(0, 20, value=6, step=1, label="Context switches")
+
+                    gr.Markdown("### 🧘 Focus & Recovery")
+                    deep_work_blocks = gr.Slider(0, 4, value=1, step=1, label="Deep work blocks (≥60 min)")
+                    break_minutes = gr.Slider(0, 120, value=30, step=5, label="Total break minutes")
+
+                    gr.Markdown("### ⏰ Workday Timing")
+                    day_start_hour = gr.Slider(6, 11, value=9, step=1, label="Start hour (24h)")
+                    day_end_hour = gr.Slider(14, 21, value=17, step=1, label="End hour (24h)")
+
+                    btn = gr.Button("🔍 Analyze Workload", variant="primary")
+
+                # OIKEA sarake: tulos
+                with gr.Column(scale=1):
+                    gr.Markdown("### 📈 Result")
+                    headline_out = gr.HTML()
+                    probs_out = gr.Label(label="Workload Probabilities")
+                    explanation_out = gr.HTML()
+
+            # Kun valitaan esimerkki → täytetään sliderit
+            example_dropdown.change(
+                load_example,
+                inputs=example_dropdown,
+                outputs=[
+                    meetings_count,
+                    total_meeting_hours,
+                    context_switches,
+                    deep_work_blocks,
+                    break_minutes,
+                    day_start_hour,
+                    day_end_hour,
+                ],
+            )
+
+            # Varsinainen ennustekutsu
+            btn.click(
+                predict_workload,
+                inputs=[
+                    meetings_count,
+                    total_meeting_hours,
+                    context_switches,
+                    deep_work_blocks,
+                    break_minutes,
+                    day_start_hour,
+                    day_end_hour,
+                ],
+                outputs=[headline_out, probs_out, explanation_out],
+            )
+
+        with gr.Tab("ℹ️ About"):
+            gr.Markdown("""
+            ### About This Tool
+
+            This demo shows how **calendar metadata** can be used to estimate cognitive workload — helping you reflect on sustainability, focus, and recovery.
+
+            - **Synthetic data only**: No real user data was used.
+            - **Model**: Trained `RandomForestClassifier` (scikit-learn).
+            - **Output**: 3-class workload (`Low`, `Medium`, `High`).
+            - **Goal**: Spark reflection, not replace judgment.
+
+            Use this to **simulate "what-if" scenarios**:  
+            _"What if I cancel two meetings?"_ or _"What if I add a 90-min focus block?"_
+
+            Built with **Python**, **scikit-learn** and **Gradio**, deployed on **Hugging Face Spaces**.
+            """)
+
+if __name__ == "__main__":
+    demo.launch()
+
+
+
+

train_model.py

@@ -0,0 +1,67 @@
+import numpy as np
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.model_selection import train_test_split
+import joblib
+
+def generate_synthetic_data(n_samples=5000, random_state=42):
+    rng = np.random.RandomState(random_state)
+
+    meetings_count = rng.randint(0, 12, size=n_samples)               # kpl
+    total_meeting_hours = rng.uniform(0, 9, size=n_samples)           # h
+    context_switches = rng.randint(0, 20, size=n_samples)             # kpl
+    deep_work_blocks = rng.randint(0, 5, size=n_samples)              # kpl
+    break_minutes = rng.randint(0, 120, size=n_samples)               # min
+    day_start_hour = rng.randint(7, 11, size=n_samples)               # 7–10
+    day_end_hour = rng.randint(14, 21, size=n_samples)                # 14–20
+
+    df = pd.DataFrame({
+        "meetings_count": meetings_count,
+        "total_meeting_hours": total_meeting_hours,
+        "context_switches": context_switches,
+        "deep_work_blocks": deep_work_blocks,
+        "break_minutes": break_minutes,
+        "day_start_hour": day_start_hour,
+        "day_end_hour": day_end_hour,
+    })
+
+    # Heuristic "actual workload" [0, 1]
+    day_length = day_end_hour - day_start_hour
+    load_score = (
+        0.3 * (meetings_count / 10)
+        + 0.25 * (total_meeting_hours / 8)
+        + 0.2 * (context_switches / 20)
+        + 0.15 * (day_length / 12)
+        - 0.15 * (deep_work_blocks / 4)
+        - 0.1 * (break_minutes / 120)
+        + rng.normal(0, 0.05, size=n_samples)
+    )
+
+    load_score = np.clip(load_score, 0, 1)
+
+    # Discretize into classes 0 = low, 1 = medium, 2 = high
+    labels = np.zeros(n_samples, dtype=int)
+    labels[load_score > 0.33] = 1
+    labels[load_score > 0.66] = 2
+
+    return df, labels
+
+if __name__ == "__main__":
+    X, y = generate_synthetic_data()
+
+    X_train, X_test, y_train, y_test = train_test_split(
+        X, y, test_size=0.2, random_state=42, stratify=y
+    )
+
+    clf = RandomForestClassifier(
+        n_estimators=150,
+        max_depth=8,
+        random_state=42
+    )
+    clf.fit(X_train, y_train)
+
+    acc = clf.score(X_test, y_test)
+    print(f"Test accuracy: {acc:.3f}")
+
+    joblib.dump(clf, "workload_model.joblib")
+    print("Saved model to workload_model.joblib")

workload_model.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8438e2fd0d4212f6fdfab15c3f60436ccf29fd503b407a87b3fd32968c0c09c9
+size 5246049