python-calc-adapter

WaveApps Python Calculation Adapter Example

This example defines a Python calculation adapter that sends data to and from the WaveApps host application.

Key User Tasks

1. Modify data_proxy.py, __init__ function, to define outputs and add any custom configuration parameters for calculation algorithm (these will be synchronized to WaveApps C# proxy adapter).
2. Modify process_data to implement desired calculation algorithm.

Usage

python main.py <hostname> <wave_apps_port> <data_pub_port>

Example

# Connect to a WaveApps proxy adapter running on localhost port 7198 and
# host data publication on port 7199. These ports must be unique for each
# Python calculation adapter.
python main.py localhost 7198 7199

The example python calculation will:

1. Connect to the specified WaveApps adapter (<hostname>:<wave_apps_port>)
2. Establish a publisher for calculated data (<data_pub_port>)
3. Subscribe to frequency measurements using filter: FILTER ActiveMeasurements WHERE SignalType = 'FREQ'
4. Receive incoming measurements, grouped by timestamp to the nearest second
5. Process grouped data in one-second windows
6. Calculate and display average frequency for each second of data
7. Trigger an event if frequency is out of range

Key Features

• Timestamp Grouping: Groups measurements by whole seconds with subsecond alignment
• Data Buffering: Maintains 5-second measurement windows for processing
• Time Validation: Filters data based on configurable lag time (2 seconds) and lead time (2 seconds)
• Downsampling Detection: Tracks and reports when data rate exceeds processing capacity
• Thread-Safe Processing: Uses locks to ensure data integrity in multi-threaded environment

Configuration

This example is configured with:

• Measurement window size: 5 seconds
• Samples per second: 3000
• Lag time: 2.0 seconds
• Lead time: 2.0 seconds

What to Expect

When connected to a publisher, you should see:

• "Measurement reader established" message
• "Receiving measurements..." when data starts arriving
• Average frequency calculations for each second of data

Example output:

Receiving measurements...

Average frequency for 3000 values in second 45: 60.001234 Hz

Average frequency for 3000 values in second 46: 59.998765 Hz

Data Processing

The example includes a process_data callback function that:

• Receives one-second buffers of time-aligned measurements
• Filters frequency values to reasonable range (59.95 to 60.05 Hz)
• Calculates average frequency across all valid measurements
• Demonstrates how to access measurement values, timestamps, and signal IDs
• Shows how to iterate through grouped measurement data

Stopping the Calculation

Press Enter to gracefully disconnect and shut down the subscriber.