npcsh 0.3.32__py3-none-any.whl → 1.0.1__py3-none-any.whl

npcsh/load_data.py

@@ -1,83 +0,0 @@
-import fitz  # PyMuPDF
-import pandas as pd
-import json
-import io
-from PIL import Image
-import numpy as np
-from typing import Optional
-
-
-def load_csv(file_path):
-    df = pd.read_csv(file_path)
-    return df
-
-
-def load_json(file_path):
-    with open(file_path, "r") as f:
-        data = json.load(f)
-    df = pd.DataFrame(data)
-    return df
-
-
-def load_txt(file_path):
-    with open(file_path, "r") as f:
-        text = f.read()
-    df = pd.DataFrame({"text": [text]})
-    return df
-
-
-def load_excel(file_path):
-    df = pd.read_excel(file_path)
-    return df
-
-
-def load_image(file_path):
-    img = Image.open(file_path)
-    img_array = np.array(img)
-    df = pd.DataFrame(
-        {
-            "image_array": [img_array.tobytes()],
-            "shape": [img_array.shape],
-            "dtype": [img_array.dtype.str],
-        }
-    )
-    return df
-
-
-def load_pdf(file_path):
-    pdf_document = fitz.open(file_path)
-    texts = []
-    images = []
-
-    for page_num, page in enumerate(pdf_document):
-        # Extract text
-        text = page.get_text()
-        texts.append({"page": page_num + 1, "content": text})
-
-        # Extract images
-        image_list = page.get_images(full=True)
-        for img_index, img in enumerate(image_list):
-            xref = img[0]
-            base_image = pdf_document.extract_image(xref)
-            image_bytes = base_image["image"]
-
-            # Convert image to numpy array
-            image = Image.open(io.BytesIO(image_bytes))
-            img_array = np.array(image)
-
-            images.append(
-                {
-                    "page": page_num + 1,
-                    "index": img_index + 1,
-                    "array": img_array.tobytes(),
-                    "shape": img_array.shape,
-                    "dtype": str(img_array.dtype),
-                }
-            )
-
-    # Create DataFrame
-    df = pd.DataFrame(
-        {"texts": json.dumps(texts), "images": json.dumps(images)}, index=[0]
-    )
-
-    return df

npcsh/main.py

@@ -1,5 +0,0 @@
-def main():
-    print("Hello from npcsh!")
-
-if __name__ == "__main__":
-    main()

npcsh/model_runner.py

@@ -1,189 +0,0 @@
-import pandas as pd
-import sqlite3
-from datetime import datetime
-
-
-class NPCModelRunner:
-    def __init__(self, npc_compiler: NPCCompiler):
-        self.compiler = npc_compiler
-        self.sql_ops = NPCSQLOperations(npc_compiler)
-        self.history_db = os.path.expanduser("~/npcsh_history.db")
-        self.model_registry = {}
-
-    def _init_history_db(self):
-        with sqlite3.connect(self.history_db) as conn:
-            conn.execute(
-                """
-                CREATE TABLE IF NOT EXISTS model_runs (
-                    model_name TEXT,
-                    run_timestamp DATETIME,
-                    run_status TEXT,
-                    metadata TEXT,
-                    output_preview TEXT
-                )
-            """
-            )
-
-            conn.execute(
-                """
-                CREATE TABLE IF NOT EXISTS model_dependencies (
-                    model_name TEXT,
-                    depends_on TEXT,
-                    created_at DATETIME
-                )
-            """
-            )
-
-    def _load_sample_data(self):
-        # For testing purposes
-        return pd.DataFrame(
-            {
-                "customer_id": range(1, 4),
-                "feedback_text": [
-                    "Great service but expensive",
-                    "Product needs improvement",
-                    "Amazing experience overall",
-                ],
-                "customer_segment": ["premium", "basic", "premium"],
-            }
-        )
-
-    def run_model(self, model_name: str, model_sql: str, sample_data=None):
-        try:
-            # Load or use sample data
-            if sample_data is None:
-                raw_data = self._load_sample_data()
-            else:
-                raw_data = sample_data
-
-            # Register the raw data as a table
-            self.model_registry["raw_customer_feedback"] = raw_data
-
-            # Parse SQL and execute with our custom functions
-            if "{{ ref(" in model_sql:
-                # Handle model dependencies
-                for dep in self._extract_dependencies(model_sql):
-                    if dep not in self.model_registry:
-                        raise ValueError(f"Dependent model {dep} not found")
-
-                # Replace ref() calls with actual data
-                model_sql = self._resolve_refs(model_sql)
-
-            # Execute the model
-            result_df = self._execute_model(model_sql)
-
-            # Store in registry
-            self.model_registry[model_name] = result_df
-
-            # Log to history
-            self._log_model_run(model_name, result_df)
-
-            return result_df
-
-        except Exception as e:
-            self._log_model_run(model_name, None, status="failed", error=str(e))
-            raise e
-
-    def _execute_model(self, model_sql: str) -> pd.DataFrame:
-        # This is a simplified version - you'd want to properly parse the SQL
-        # For now, let's implement basic synthesize functionality
-
-        if "synthesize(" in model_sql:
-            raw_data = self.model_registry["raw_customer_feedback"]
-            return self.sql_ops.synthesize(
-                df=raw_data,
-                column="feedback_text",
-                npc="analyst",
-                context="customer_segment",
-                framework="satisfaction",
-            )
-
-        if "spread_and_sync(" in model_sql:
-            input_data = self.model_registry["customer_feedback"]
-            return self.sql_ops.spread_and_sync(
-                df=input_data,
-                column="feedback_analysis",
-                npc="strategy_agent",
-                variations=["short_term", "long_term"],
-                sync_strategy="balanced_analysis",
-                context=self._load_vars()["business_context"],
-            )
-
-        return pd.DataFrame()  # Fallback
-
-    def _log_model_run(self, model_name: str, result_df, status="success", error=None):
-        with sqlite3.connect(self.history_db) as conn:
-            metadata = {
-                "status": status,
-                "error": error,
-                "rows_processed": len(result_df) if result_df is not None else 0,
-            }
-
-            preview = result_df.head().to_dict() if result_df is not None else None
-
-            conn.execute(
-                """INSERT INTO model_runs
-                   (model_name, run_timestamp, run_status, metadata, output_preview)
-                   VALUES (?, ?, ?, ?, ?)""",
-                (
-                    model_name,
-                    datetime.now().isoformat(),
-                    status,
-                    json.dumps(metadata),
-                    json.dumps(preview),
-                ),
-            )
-
-    def _load_vars(self):
-        with open("npc_project.yml", "r") as f:
-            return yaml.safe_load(f).get("vars", {})
-
-    def _extract_dependencies(self, model_sql: str) -> List[str]:
-        # Simple regex to extract model names from ref() calls
-        import re
-
-        refs = re.findall(r'{{\s*ref\([\'"](.+?)[\'"]\)\s*}}', model_sql)
-        return refs
-
-    def _resolve_refs(self, model_sql: str) -> str:
-        # Replace ref() calls with actual table references
-        import re
-
-        def replace_ref(match):
-            model_name = match.group(1)
-            if model_name in self.model_registry:
-                return f"model_{model_name}"
-            raise ValueError(f"Model {model_name} not found")
-
-        return re.sub(r'{{\s*ref\([\'"](.+?)[\'"]\)\s*}}', replace_ref, model_sql)
-
-
-# Usage example:
-def main():
-    # Initialize
-    npc_compiler = NPCCompiler("~/.npcsh/npc_team", "~/npcsh_history.db")
-    runner = NPCModelRunner(npc_compiler)
-
-    # Run first model
-    with open("models/customer_feedback.sql", "r") as f:
-        feedback_model = runner.run_model("customer_feedback", f.read())
-        print("First model results:")
-        print(feedback_model.head())
-
-    # Run second model that depends on the first
-    with open("models/customer_insights.sql", "r") as f:
-        insights_model = runner.run_model("customer_insights", f.read())
-        print("\nSecond model results:")
-        print(insights_model.head())
-
-    # Check history
-    with sqlite3.connect(runner.history_db) as conn:
-        history = pd.read_sql(
-            "SELECT * FROM model_runs ORDER BY run_timestamp DESC", conn
-        )
-        print("\nModel run history:")
-        print(history)
-
-
-if __name__ == "__main__":
-    main()