Working with correct csv path but no output

Prompt_engineer/config.yaml

@@ -0,0 +1,12 @@
+# Configuration file for the medical report generator
+cleaned_data_csv_path: '/Users/ayodejioyesanya/Desktop/Tdata/cleaned_medical_data.csv'
+prompt_template: |
+  <s>[INST] <<SYS>> You are an experienced medical AI assistant trained to provide helpful and accurate 
+  information to patients. You have extensive knowledge of human anatomy, common medical conditions, and 
+  evidence-based treatments. Your responses should be empathetic, informative, and adhere to medical best 
+  practices. You will not provide any medical advice that could be harmful. [/INST]
+
+  Diagnosis: {diagnosis}. Patient's age: {age}. Gender: {gender}. 
+  Conditions: {conditions}. Observations: {observations}. Care plans: {care_plans}. 
+  Imaging modality: {modality}. Body area: {body_area}. 
+  Please provide a comprehensive report based on the patient's data and imaging results.</s>

Prompt_engineer/main.py

@@ -0,0 +1,82 @@
+import argparse
+import yaml
+import pandas as pd
+from datetime import datetime
+from typing import Dict, Any, Optional
+
+def load_config(config_path: str) -> Dict[str, Any]:
+    """
+    Load the YAML configuration file.
+    """
+    with open(config_path, 'r') as file:
+        config = yaml.safe_load(file) or {}
+    return config
+
+def calculate_age(birthdate: str) -> int:
+    """
+    Calculate age given the birthdate.
+    """
+    birth_date = datetime.strptime(birthdate, "%Y-%m-%d")
+    today = datetime.now()
+    return today.year - birth_date.year - ((today.month, today.day) < (birth_date.month, birth_date.day))
+
+def generate_medical_prompt(patient_data: Dict[str, Any], diagnosis: str, config: Dict[str, Any]) -> str:
+    """
+    Generate a medical prompt for the medical report based on patient data and diagnosis details using a template from the configuration.
+    Uses REASONDESCRIPTION as the source for the patient's conditions.
+    """
+    prompt_template = config.get('prompt_template', "Default template if not specified in config.")
+    age = calculate_age(patient_data['BIRTHDATE'])  # Calculating age using the BIRTHDATE field
+    conditions = patient_data.get('REASONDESCRIPTION', 'No reason description provided')  # Default message if not specified
+    prompt = prompt_template.format(
+        diagnosis=diagnosis,
+        age=age,
+        gender=patient_data['GENDER'],
+        conditions=conditions,  # Correct field used
+        observations=patient_data.get('observation', 'No observations recorded'),  # Use correct field for observations
+        care_plans=patient_data.get('DESCRIPTION_careplan', 'No care plans recorded'),
+        modality=patient_data.get('modality', "Not specified"),
+        body_area=patient_data.get('body_area', "Not specified")
+    )
+    return prompt
+
+def load_patient_data(csv_path: str) -> pd.DataFrame:
+    """
+    Load patient data from a CSV file.
+    """
+    return pd.read_csv(csv_path)
+
+def select_random_patient_data(patient_data: pd.DataFrame) -> Dict[str, Any]:
+    """
+    Select random patient data from the entire dataset, independent of the diagnosis.
+    """
+    return patient_data.sample(n=1).iloc[0].to_dict()
+
+def parse_arguments() -> argparse.Namespace:
+    """
+    Parse command line arguments.
+    """
+    parser = argparse.ArgumentParser(description='Medical Report Generator')
+    parser.add_argument('--config_path', type=str, default='/Users/ayodejioyesanya/Documents/SFdev/Prompt_engineer/config.yaml', help='Path to the YAML configuration file.')
+    parser.add_argument('--csv_path', type=str, default='/Users/ayodejioyesanya/Desktop/Tdata/cleaned_medical_data.csv', help='Path to the CSV file containing patient data.')
+    parser.add_argument('--diagnosis', type=str, required=True, help='Diagnosis determined by the image classifier.')
+    return parser.parse_args()
+
+def main() -> None:
+    """
+    Main function to orchestrate the workflow.
+    """
+    args = parse_arguments()
+    config = load_config(args.config_path)
+    patient_data = load_patient_data(args.csv_path)
+    diagnosis = args.diagnosis  # Diagnosis passed from the command line
+    random_patient_data = select_random_patient_data(patient_data)
+
+    if random_patient_data:
+        medical_prompt = generate_medical_prompt(random_patient_data, diagnosis, config)
+        print(medical_prompt)
+    else:
+        print("No patient data available.")
+
+if __name__ == '__main__':
+    main()

Prompt_engineer/synthea_data.py

@@ -0,0 +1,66 @@
+from typing import Dict, Any, Tuple, Optional
+from datetime import datetime
+from dateutil.relativedelta import relativedelta
+import random
+
+# Example static data, simulating a patient database
+PATIENT_DATA = [
+    {
+        'BIRTHDATE': '1990-01-01',
+        'GENDER': 'Male',
+        'REASONDESCRIPTION': 'lung cancer',
+        'observation': 'increased cough',
+        'DESCRIPTION_careplan': 'regular monitoring',
+        'modality': 'X-Ray',
+        'body_area': 'Chest'
+    },
+    {
+        'BIRTHDATE': '1985-05-15',
+        'GENDER': 'Female',
+        'REASONDESCRIPTION': '',  # This patient has no diagnosis specified.
+        'observation': 'shortness of breath',
+        'DESCRIPTION_careplan': 'oxygen therapy',
+        'modality': 'CT Scan',
+        'body_area': 'Chest'
+    },
+    # Additional records can be added here.
+]
+
+class SyntheaData:
+    """
+    Simulates data retrieval from a static dataset.
+    """
+    def get_patient_data_by_diagnosis(self, diagnosis: str) -> Tuple[Dict[str, Any], Optional[str], Optional[str]]:
+        """
+        Retrieves a random patient's data who has been diagnosed with a specified condition using the REASONDESCRIPTION.
+        If no diagnosis is provided in REASONDESCRIPTION, it defaults to "Nil significant past medical history."
+        """
+        # Filtering patients based on diagnosis, considering those with empty or null diagnosis as having no significant history.
+        matching_patients = [patient for patient in PATIENT_DATA if patient['REASONDESCRIPTION'].strip().lower() == diagnosis.lower()] if diagnosis.strip() else [patient for patient in PATIENT_DATA if not patient['REASONDESCRIPTION'].strip()]
+
+        if matching_patients:
+            selected_patient = random.choice(matching_patients)
+            age = relativedelta(datetime.now(), datetime.strptime(selected_patient['BIRTHDATE'], '%Y-%m-%d')).years
+            patient_data = {
+                'age': age,
+                'gender': selected_patient['GENDER'],
+                'conditions': selected_patient['REASONDESCRIPTION'] if selected_patient['REASONDESCRIPTION'].strip() else "Nil significant past medical history",
+                'observations': selected_patient['observation'],
+                'care_plans': selected_patient['DESCRIPTION_careplan'],
+                'modality': selected_patient.get('modality', "Not specified"),
+                'body_area': selected_patient.get('body_area', "Not specified")
+            }
+            modality = selected_patient.get('modality')
+            body_area = selected_patient.get('body_area')
+            return patient_data, modality, body_area
+        else:
+            # Default case when no patients match the criteria, including empty diagnosis search
+            return ({
+                'age': None,
+                'gender': None,
+                'conditions': "Nil significant past medical history",
+                'observations': None,
+                'care_plans': None,
+                'modality': None,
+                'body_area': None
+            }, None, None)