scitex-db 0.1.4

Database utilities for scientific computing with SQLite3 and PostgreSQL

scitex-db

Database utilities for scientific computing.

Interfaces: Python ⭐⭐⭐ (primary) · CLI ⭐ · MCP — · Skills ⭐⭐ · Hook — · HTTP —

Problem and Solution

#	Problem	Solution
1	Storing ndarrays in SQLite means pickle.dumps → BLOB -- no compression, no type info, no deterministic hashing	SQLite3.save_array(name, arr) / load_array(name) -- compressed BLOB storage with typed round-trip; compatible with pandas via to_df
2	sqlite3 API is low-level -- every project re-writes connect/transaction/execute boilerplate	with db: context-manager transactions -- health checks, duplicate removal, schema inspection built in

Overview

scitex-db provides enhanced database operations designed for scientific research:

Features

SQLite3 with Scientific Extensions:

• 📊 Array Storage - Store/retrieve NumPy arrays efficiently
• 🔬 Blob Storage - Serialize Python objects with metadata
• 📦 Batch Operations - High-performance bulk inserts
• 🔍 Advanced Queries - Scientific query patterns
• 🗂️ Git Integration - Version control for databases
• 📤 Import/Export - CSV, JSON, DataFrame conversions
• 🔧 Maintenance Tools - Health checks, deduplication

PostgreSQL Support:

• Full-featured PostgreSQL wrapper
• Optimized for scientific datasets

CLI Tools:

scitex-db inspect database.db
scitex-db health database.db --fix

Installation

pip install scitex-db

For PostgreSQL:

pip install scitex-db[postgresql]

For all features:

pip install scitex-db[all]

Quick Start

Basic Usage

from scitex_db import SQLite3

# Initialize
db = SQLite3("experiments.db")

# Create table
db.create_table("results", {
    "id": "INTEGER PRIMARY KEY",
    "experiment": "TEXT",
    "accuracy": "REAL"
})

# Insert data
db.insert_many("results", [
    {"experiment": "exp1", "accuracy": 0.95},
    {"experiment": "exp2", "accuracy": 0.92}
])

# Query
results = db.get_rows("results", where="accuracy > 0.9")
print(results)

Array Storage

import numpy as np

# Save arrays
data = np.random.rand(1000, 50)
db.save_array("features", data,
              column="embeddings",
              additional_columns={"model": "bert"})

# Load arrays
loaded = db.load_array("features", "embeddings",
                       where="model = 'bert'")

Blob Storage

# Store arbitrary objects
model = {"weights": np.random.rand(100), "config": {...}}
db.save_blob("models", model,
             column="checkpoint",
             additional_columns={"epoch": 10})

# Retrieve
model = db.load_blob("models", "checkpoint", where="epoch = 10")

Git Integration

from scitex_db import SQLite3

db = SQLite3("versioned.db")
db.init_git()  # Initialize git tracking

# Automatic commits on changes
db.insert("results", {"value": 42})
# Commits with message: "Insert 1 row(s) into results"

Advanced Features

Transaction Management

with db.transaction():
    db.insert("table1", {...})
    db.insert("table2", {...})
    # Auto-commit on success, rollback on error

Batch Operations

# High-performance bulk insert
large_dataset = [{"id": i, "value": i**2} for i in range(10000)]
db.insert_many("data", large_dataset, batch_size=1000)

Database Inspection

# Get comprehensive summary
db.summary  # or db()

# Inspect specific table
db.inspect_table("results")

# Health check
from scitex_db import check_health
check_health("database.db", fix_issues=True)

Part of SciTeX Ecosystem

• scitex-core - Core infrastructure
• scitex-io - Data I/O (can use scitex-db)
• scitex-writer - Academic writing
• scitex-scholar - Paper management
• scitex - Main package

License

MIT License - see LICENSE file for details.