Variable Backup & Restore¶

Every Strategy object has a vars attribute (self.vars), an instance of the Vars class. It stores runtime variables and is periodically backed up to the database specified by the DB_CONNECTION_STR environment variable.

How It Works¶

Loading Variables: Before the first trading iteration, saved variables are loaded into self.vars.
Backing Up Variables: After each trading iteration and when the bot stops or crashes, self.vars is automatically backed up to the database.

Benefits of self.vars¶

Persistence: Resume from the last state after interruptions.
Namespacing: Prevent conflicts with other components or strategies.
Scalability: Efficient storage, especially when scaling or using multiple strategies.

Usage Guide¶

Setting Variables:

Assign values using attribute notation:

def on_trading_iteration(self):
    self.vars.trade_count = self.vars.get('trade_count', 0) + 1

Accessing Variables:

Retrieve variables using attribute notation:

def on_trading_iteration(self):
    current_count = self.vars.trade_count
    print(f"Current trade count: {current_count}")

Getting All Variables:

Use the all() method:

all_variables = self.vars.all()
print(all_variables)

# Output: {'trade_count': 5, 'last_price': 102.5}

Tip

Use all() for debugging or processing multiple variables at once.

Database Configuration¶

Set the DB_CONNECTION_STR environment variable in your .env file:

DB_CONNECTION_STR="postgresql://user:password@localhost:5432/database_name"

Database Storage Structure¶

Variables are stored in a PostgreSQL table:

Table Name: Defined by self.backup_table_name (default is vars_backup).
Columns: - id (Primary Key) - last_updated: Timestamp of the last backup. - variables: JSON string of the strategy’s variables (self.vars.all()). - strategy_id: Strategy name (self.name).

Example Database:

id	last_updated	variables	strategy_id
550e8400	2023-10-05 14:30:00	{“var1”: 10, “var2”: “A”}	OptionsCondorMartingale