Dip-SIP Local (Streamlit)

Local-only or cloud-hosted Streamlit application for:

• Index selection (NIFTY 50 / Next 50 / IT / Pharma, extendable)
• Data ingestion and caching (CSV upload + pluggable providers)
• Dip-SIP trigger generation (daily band-entry) + cash bucket tracking
• Backtest comparison vs Standard SIP (terminal value + XIRR)
• Exports (ledger CSV + summary JSON)
• User authentication (login/logout)
• Supabase storage (persistent online) or SQLite (local-only)

---

Quick start (local — Mac/Linux)

git clone -b dev https://github.com/v-bansal/dip_sip.git
cd dip_sip
python3 -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt

# Copy and configure secrets (optional for local SQLite)
cp .streamlit/secrets.toml.example .streamlit/secrets.toml
# Edit .streamlit/secrets.toml and set auth_cookie_key

streamlit run app.py

Default login (change in config/credentials.yaml):

• Username: admin | Password: admin123
• Username: viewer | Password: viewer123

---

Quick start (online — Streamlit Community Cloud)

1. Create Supabase project

1. Go to supabase.com → New project
2. Open SQL Editor → paste storage/supabase_schema.sql → Run
3. Go to Settings → API → copy URL and anon public key

2. Deploy to Streamlit Cloud

1. Go to share.streamlit.io
2. New app → Select v-bansal/dip_sip, branch dev, file app.py
3. Click Advanced settings → Secrets → paste:

   auth_cookie_name = "dip_sip_auth"
   auth_cookie_key = "your_random_32char_key"  # Generate: python -c "import secrets; print(secrets.token_hex(16))"
   auth_cookie_expiry_days = 30
   SUPABASE_URL = "https://xxx.supabase.co"
   SUPABASE_KEY = "your-anon-key"

4. Deploy → Your app is live with authentication + persistent storage

---

Data input

The app expects daily data with:

• Date (YYYY-MM-DD)
• Close (index level; TRI preferred)

Use the Data Manager page to upload CSV and save it into the cache.

---

Architecture

Local mode (SQLite)

• Runs on your Mac/laptop
• Data stored in ./data/cache.sqlite
• No internet needed after setup

Online mode (Supabase)

• Runs on Streamlit Cloud
• Data stored in Supabase Postgres (persistent across deploys)
• Access from anywhere
• Multi-user capable

Auto-detection

The app automatically uses:

• Supabase if SUPABASE_URL exists in .streamlit/secrets.toml
• SQLite otherwise (local development)

---

Authentication

Change passwords

Edit config/credentials.yaml. Generate bcrypt hashes:

python -c "import bcrypt; print(bcrypt.hashpw('yourpassword'.encode(), bcrypt.gensalt()).decode())"

Add users

Add to config/credentials.yaml:

credentials:
  usernames:
    newuser:
      email: [email protected]
      name: New User
      password: $2b$12$...  # bcrypt hash

---

Providers (data sources)

• ✅ CSV upload — works offline (implemented)
• 🚧 NIFTY Indices download — stub (implement in providers/niftyindices.py)
• 🚧 NSE Historical Index — stub (implement in providers/nse.py)
• 🚧 SmartAPI / Breeze — stubs (for broker APIs)

---

Local cache tables

SQLite (local)

• prices (normalized daily series)
• runs (backtest runs + parameters)
• ledgers (day-by-day ledger for a run)

Supabase (online)

Same schema, created via storage/supabase_schema.sql

---

Files

Path	What it is
app.py	Main dashboard with auth
pages/01_Data_Manager.py	CSV upload → cache
pages/02_Run_Viewer.py	Saved run viewer
core/engine.py	Backtest engine
storage/cache.py	SQLite adapter
storage/supabase_cache.py	Supabase adapter
storage/cache_factory.py	Auto-selects SQLite or Supabase
config/credentials.yaml	User logins (bcrypt hashed)
config/index_registry.yaml	Index list
config/defaults.yaml	Strategy defaults

---

Troubleshooting

Local

Problem	Fix
ModuleNotFoundError: streamlit_authenticator	Run pip install -r requirements.txt
KeyError: 'auth_cookie_key'	Create .streamlit/secrets.toml from example
Login fails	Check config/credentials.yaml passwords

Online (Streamlit Cloud)

Problem	Fix
Unable to connect to Supabase	Check SUPABASE_URL and SUPABASE_KEY in secrets
Tables don't exist	Run storage/supabase_schema.sql in Supabase SQL Editor
Data not persisting	Confirm you're using Supabase, not SQLite

---

Next steps

1. Change default passwords in config/credentials.yaml
2. Deploy to Streamlit Cloud with Supabase
3. Implement auto-fetch providers (niftyindices.py or nse.py)
4. Add more indices in config/index_registry.yaml
5. Customize strategy defaults in config/defaults.yaml