Update app.py

app.py

@@ -1,160 +1,192 @@
 import gradio as gr
-  import pandas as pd
-  from fastapi import FastAPI, Request
-  from fastapi.responses import JSONResponse
-  import uvicorn
-  from sklearn.linear_model import LinearRegression
-  from sklearn.model_selection import train_test_split
-  import os
-  import uuid
-  from datetime import datetime
-  from fastapi.staticfiles import StaticFiles
-  from reportlab.lib.pagesizes import A4
-  from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer
-  from reportlab.lib.styles import getSampleStyleSheet
-
-  # Hardcoded mappings
-  weather_map = {"Cloudy": 0, "Rainy": 1, "Sunny": 2}
-
-  # Load and preprocess training data
-  df = pd.read_csv("new_delay_data.csv")
-
-  # Encode categorical features
-  df = pd.get_dummies(df, columns=["Phase"], drop_first=True)  # Finishing as baseline
-  df["Weather"] = df["Weather"].map(weather_map)
-
-  # Handle missing or invalid mappings
-  df.dropna(subset=["Phase_Framing", "Phase_Foundation", "Weather", "Absentee", "DelayLog", "Delay%"], inplace=True)
-
-  # Split features and target
-  X = df[["Phase_Framing", "Phase_Foundation", "Weather", "Absentee", "DelayLog"]]
-  y = df["Delay%"]
-
-  # Train model
-  model = LinearRegression()
-  model.fit(X, y)
-
-  # Function to clean up old files
-  def clean_tmp():
-      directory = "./static_files"
-      for file in os.listdir(directory):
-          if file.endswith(".pdf"):
-              try:
-                  os.remove(os.path.join(directory, file))
-              except OSError:
-                  pass
-
-  # Function to generate PDF using reportlab
-  def generate_pdf(phase, weather, absentee_pct, delay_log, prediction, risk, insight):
-      os.makedirs("./static_files", exist_ok=True)
-      clean_tmp()
-      pdf_filename = f"delay_report_{uuid.uuid4().hex}.pdf"
-      pdf_path = f"./static_files/{pdf_filename}"
-      
-      doc = SimpleDocTemplate(pdf_path, pagesize=A4)
-      styles = getSampleStyleSheet()
-      story = []
-
-      story.append(Paragraph("Delay Prediction Report", styles['Title']))
-      story.append(Spacer(1, 12))
-      story.append(Paragraph(f"Generated on: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}", styles['Normal']))
-      story.append(Spacer(1, 12))
-      
-      story.append(Paragraph("Project Details", styles['Heading2']))
-      story.append(Paragraph(f"Phase: {phase}", styles['Normal']))
-      story.append(Paragraph(f"Weather: {weather}", styles['Normal']))
-      story.append(Paragraph(f"Absentee %: {absentee_pct:.2f}%", styles['Normal']))
-      story.append(Paragraph(f"Previous Delay Log: {delay_log:.2f} days", styles['Normal']))
-      story.append(Spacer(1, 12))
-      
-      story.append(Paragraph("Prediction Results", styles['Heading2']))
-      story.append(Paragraph(f"Predicted Delay: {prediction:.2f}%", styles['Normal']))
-      story.append(Paragraph(f"Risk Level: {risk}", styles['Normal']))
-      story.append(Paragraph(f"AI Insight: {insight}", styles['Normal']))
-      
-      doc.build(story)
-      
-      pdf_url = f"https://ajaykumarpilla-delay-predictor.hf.space/static/{pdf_filename}"
-      return pdf_url, pdf_path
-
-  # Main prediction function
-  def predict_delay(phase, weather, absentee_pct, delay_log):
-      framing = 1 if phase == "Framing" else 0
-      foundation = 1 if phase == "Foundation" else 0
-      weather_encoded = weather_map.get(weather, 0)
-      input_data = [[framing, foundation, weather_encoded, absentee_pct, delay_log]]
-      
-      prediction = model.predict(input_data)[0]
-      prediction = round(prediction, 2)
-
-      if prediction >= 75:
-          risk = "High Risk"
-      elif prediction >= 50:
-          risk = "Moderate Risk"
-      else:
-          risk = "Low Risk"
-
-      insight = f"Phase: {phase}, Weather: {weather}, Absenteeism: {absentee_pct}%, Previous Delay: {delay_log} → Risk: {risk}"
-
-      pdf_url, pdf_path = generate_pdf(phase, weather, absentee_pct, delay_log, prediction, risk, insight)
-      return prediction, risk, insight, pdf_url
-
-  # FastAPI for Salesforce
-  api_app = FastAPI()
-  os.makedirs("./static_files", exist_ok=True)
-  api_app.mount("/static", StaticFiles(directory="./static_files"), name="static")
-
-  @api_app.post("/predict")
-  async def predict_from_salesforce(request: Request):
-      try:
-          data = await request.json()
-          phase = data.get("phase", "Framing")
-          weather = data.get("weather", "Sunny")
-          absentee_pct = float(data.get("absentee_pct", 0))
-          delay_log = float(data.get("delay_log", 0))
-
-          prediction, risk, insight, pdf_url = predict_delay(phase, weather, absentee_pct, delay_log)
-
-          return JSONResponse(content={
-              "delay_probability": prediction,
-              "risk_alert": risk,
-              "ai_insight": insight,
-              "pdf_url": pdf_url,
-              "status": "success"
-          })
-      except Exception as e:
-          return JSONResponse(status_code=500, content={"status": "error", "message": str(e)})
-
-  # Gradio UI for manual testing
-  with gr.Blocks() as demo:
-      gr.Markdown("## 🏗️ Delay Predictor")
-      with gr.Row():
-          phase_input = gr.Textbox(label="Phase (Framing/Foundation/Finishing)")
-          weather_input = gr.Textbox(label="Weather (Sunny/Rainy/Cloudy)")
-      with gr.Row():
-          absentee_input = gr.Number(label="Absentee %")
-          delay_input = gr.Number(label="Previous Delay Log")
-      output = gr.Textbox(label="Prediction Summary")
-      pdf_output = gr.File(label="Download PDF Report")
-
-      def predict_and_format(phase, weather, absentee, delay_log):
-          try:
-              prediction, risk, insight, pdf_url = predict_delay(phase, weather, absentee, delay_log)
-              return (
-                  f"Predicted Delay: {prediction}%\nRisk Level: {risk}\nInsight: {insight}\nPDF URL: {pdf_url}",
-                  pdf_url
-              )
-          except Exception as e:
-              return f"Error: {str(e)}", None
-
-      submit = gr.Button("Predict")
-      submit.click(
-          predict_and_format,
-          inputs=[phase_input, weather_input, absentee_input, delay_input],
-          outputs=[output, pdf_output]
-      )
-
-  app = gr.mount_gradio_app(api_app, demo, path="/")
-
-  if __name__ == "__main__":
-      uvicorn.run(app, host="0.0.0.0", port=7860)
+import pandas as pd
+from fastapi import FastAPI, Request
+from fastapi.responses import JSONResponse
+import uvicorn
+from sklearn.linear_model import LinearRegression
+from sklearn.model_selection import train_test_split
+import os
+import subprocess
+import uuid
+from datetime import datetime
+from fastapi.staticfiles import StaticFiles
+
+# Hardcoded mappings
+weather_map = {"Cloudy": 0, "Rainy": 1, "Sunny": 2}
+
+# Load and preprocess training data
+df = pd.read_csv("new_delay_data.csv")
+
+# Encode categorical features
+df = pd.get_dummies(df, columns=["Phase"], drop_first=True)  # Finishing as baseline
+df["Weather"] = df["Weather"].map(weather_map)
+
+# Handle missing or invalid mappings
+df.dropna(subset=["Phase_Framing", "Phase_Foundation", "Weather", "Absentee", "DelayLog", "Delay%"], inplace=True)
+
+# Split features and target
+X = df[["Phase_Framing", "Phase_Foundation", "Weather", "Absentee", "DelayLog"]]
+y = df["Delay%"]
+
+# Train model
+model = LinearRegression()
+model.fit(X, y)
+
+# Function to clean up old files
+def clean_tmp():
+    directory = "./static_files"
+    for file in os.listdir(directory):
+        if file.endswith((".pdf", ".tex", ".aux", ".log")):
+            try:
+                os.remove(os.path.join(directory, file))
+            except OSError:
+                pass
+
+# Function to generate LaTeX content for PDF
+def generate_latex_content(phase, weather, absentee_pct, delay_log, prediction, risk, insight):
+    return f"""
+\\documentclass[a4paper,12pt]{{article}}
+\\usepackage{{geometry}}
+\\usepackage{{utf8x}}
+\\usepackage{{parskip}}
+\\usepackage{{titlesec}}
+\\usepackage{{xcolor}}
+\\geometry{{margin=1in}}
+\\titleformat{{\\section}}{{\\Large\\bfseries\\color{{blue}}}}{{\\thesection}}{{1em}}{{}}
+\\begin{{document}}
+\\section*{{Delay Prediction Report}}
+\\textbf{{Generated on:}} {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\\\\
+\\textbf{{Project Details}}\\\\
+\\begin{{itemize}}
+    \\item \\textbf{{Phase:}} {phase}
+    \\item \\textbf{{Weather:}} {weather}
+    \\item \\textbf{{Absentee \\%:}} {absentee_pct:.2f}\\%
+    \\item \\textbf{{Previous Delay Log:}} {delay_log:.2f} days
+\\end{{itemize}}
+\\vspace{{0.5cm}}
+\\textbf{{Prediction Results}}\\\\
+\\begin{{itemize}}
+    \\item \\textbf{{Predicted Delay:}} {prediction:.2f}\\%
+    \\item \\textbf{{Risk Level:}} {risk}
+    \\item \\textbf{{AI Insight:}} {insight}
+\\end{{itemize}}
+\\end{{document}}
+"""
+
+# Function to generate PDF and return its URL
+def generate_pdf(phase, weather, absentee_pct, delay_log, prediction, risk, insight):
+    os.makedirs("./static_files", exist_ok=True)
+    clean_tmp()  # Clean up old files
+    pdf_filename = f"delay_report_{uuid.uuid4().hex}.pdf"
+    tex_filename = f"./static_files/{pdf_filename.replace('.pdf', '.tex')}"
+    pdf_path = f"./static_files/{pdf_filename}"
+    
+    latex_content = generate_latex_content(phase, weather, absentee_pct, delay_log, prediction, risk, insight)
+    with open(tex_filename, "w") as f:
+        f.write(latex_content)
+    
+    try:
+        subprocess.run(
+            ["pdflatex", f"-output-directory=./static_files", tex_filename],
+            check=True,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            text=True
+        )
+    except subprocess.CalledProcessError as e:
+        raise Exception(f"PDF generation failed: {e.stderr}")
+    
+    for ext in [".aux", ".log", ".tex"]:
+        try:
+            os.remove(tex_filename.replace(".tex", ext))
+        except OSError:
+            pass
+    
+    pdf_url = f"https://ajaykumarpilla-delay-predictor.hf.space/static/{pdf_filename}"
+    return pdf_url, pdf_path
+
+# Main prediction function
+def predict_delay(phase, weather, absentee_pct, delay_log):
+    framing = 1 if phase == "Framing" else 0
+    foundation = 1 if phase == "Foundation" else 0
+    weather_encoded = weather_map.get(weather, 0)
+    input_data = [[framing, foundation, weather_encoded, absentee_pct, delay_log]]
+    
+    prediction = model.predict(input_data)[0]
+    prediction = round(prediction, 2)
+
+    if prediction >= 75:
+        risk = "High Risk"
+    elif prediction >= 50:
+        risk = "Moderate Risk"
+    else:
+        risk = "Low Risk"
+
+    insight = f"Phase: {phase}, Weather: {weather}, Absenteeism: {absentee_pct}%, Previous Delay: {delay_log} → Risk: {risk}"
+
+    pdf_url, pdf_path = generate_pdf(phase, weather, absentee_pct, delay_log, prediction, risk, insight)
+    return prediction, risk, insight, pdf_url
+
+# FastAPI for Salesforce
+api_app = FastAPI()
+
+# Create static_files directory at startup
+os.makedirs("./static_files", exist_ok=True)
+
+# Mount static files directory
+api_app.mount("/static", StaticFiles(directory="./static_files"), name="static")
+
+@api_app.post("/predict")
+async def predict_from_salesforce(request: Request):
+    try:
+        data = await request.json()
+        phase = data.get("phase", "Framing")
+        weather = data.get("weather", "Sunny")
+        absentee_pct = float(data.get("absentee_pct", 0))
+        delay_log = float(data.get("delay_log", 0))
+
+        prediction, risk, insight, pdf_url = predict_delay(phase, weather, absentee_pct, delay_log)
+
+        return JSONResponse(content={
+            "delay_probability": prediction,
+            "risk_alert": risk,
+            "ai_insight": insight,
+            "pdf_url": pdf_url,
+            "status": "success"
+        })
+    except Exception as e:
+        return JSONResponse(status_code=500, content={"status": "error", "message": str(e)})
+
+# Gradio UI for manual testing
+with gr.Blocks() as demo:
+    gr.Markdown("## �(:construction:) Delay Predictor")
+    with gr.Row():
+        phase_input = gr.Textbox(label="Phase (Framing/Foundation/Finishing)")
+        weather_input = gr.Textbox(label="Weather (Sunny/Rainy/Cloudy)")
+    with gr.Row():
+        absentee_input = gr.Number(label="Absentee %")
+        delay_input = gr.Number(label="Previous Delay Log")
+    output = gr.Textbox(label="Prediction Summary")
+    pdf_output = gr.File(label="Download PDF Report")
+
+    def predict_and_format(phase, weather, absentee, delay_log):
+        try:
+            prediction, risk, insight, pdf_url = predict_delay(phase, weather, absentee, delay_log)
+            return (
+                f"Predicted Delay: {prediction}%\nRisk Level: {risk}\nInsight: {insight}\nPDF URL: {pdf_url}",
+                pdf_url
+            )
+        except Exception as e:
+            return f"Error: {str(e)}", None
+
+    submit = gr.Button("Predict")
+    submit.click(
+        predict_and_format,
+        inputs=[phase_input, weather_input, absentee_input, delay_input],
+        outputs=[output, pdf_output]
+    )
+
+app = gr.mount_gradio_app(api_app, demo, path="/")
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)