⚡️ Speed up method CSVSink.parse_detection_data by 6%

[codeflash-ai]

📄 6% (0.06x) speedup for CSVSink.parse_detection_data in supervision/detection/tools/csv_sink.py

⏱️ Runtime : 366 microseconds → 345 microseconds (best of 74 runs)

📝 Explanation and details

o3-mini
We optimized the function by caching repeated attribute lookups (e.g. detections.xyxy and detections.data) and avoiding checking for the "data" attribute on every iteration (by precomputing a boolean flag). We also preallocate the result list and use local variables for faster access. The resulting code performs exactly the same functionality while reducing overhead from repeated attribute accesses.

✅ Correctness verification report:

Test	Status
⚙️ Existing Unit Tests	🔘 None Found
🌀 Generated Regression Tests	✅ 6 Passed
⏪ Replay Tests	🔘 None Found
🔎 Concolic Coverage Tests	🔘 None Found
📊 Tests Coverage	83.3%

🌀 Generated Regression Tests Details

from __future__ import annotations

import csv
from typing import Any, Dict, List, Optional

# imports
import pytest  # used for our unit tests
from supervision.detection.tools.csv_sink import CSVSink


class Detections:
    def __init__(self, xyxy, class_id=None, confidence=None, tracker_id=None, data=None):
        self.xyxy = xyxy
        self.class_id = class_id
        self.confidence = confidence
        self.tracker_id = tracker_id
        self.data = data
from supervision.detection.tools.csv_sink import CSVSink

# unit tests

# Basic Functionality






def test_empty_detections():
    detections = Detections(
        xyxy=[],
        class_id=[],
        confidence=[],
        tracker_id=[]
    )
    expected = []
    codeflash_output = CSVSink.parse_detection_data(detections)

# Large Scale Test Cases


def test_invalid_xyxy_format():
    detections = Detections(
        xyxy=[10, 20, 30, 40],  # Incorrect format
        class_id=[1],
        confidence=[0.9],
        tracker_id=[101]
    )
    with pytest.raises(TypeError):
        CSVSink.parse_detection_data(detections)

# Complex Data Structures



from __future__ import annotations

import csv
from typing import Any, Dict, List, Optional

import numpy as np
# imports
import pytest  # used for our unit tests
from supervision.detection.tools.csv_sink import CSVSink


# Mock Detections class for testing
class Detections:
    def __init__(self, xyxy, class_id=None, confidence=None, tracker_id=None, data=None):
        self.xyxy = xyxy
        self.class_id = class_id
        self.confidence = confidence
        self.tracker_id = tracker_id
        self.data = data
from supervision.detection.tools.csv_sink import CSVSink

# unit tests



def test_empty_detections():
    detections = Detections(xyxy=[])
    expected_output = []
    codeflash_output = CSVSink.parse_detection_data(detections)



def test_array_additional_data():
    detections = Detections(xyxy=[[10, 20, 30, 40]], data={"score": np.array([0.95])})
    expected_output = [{"x_min": 10, "y_min": 20, "x_max": 30, "y_max": 40, "class_id": "", "confidence": "", "tracker_id": "", "score": 0.95}]
    codeflash_output = CSVSink.parse_detection_data(detections)



def test_large_additional_data():
    detections = Detections(xyxy=[[i, i+1, i+2, i+3] for i in range(1000)], data={"score": np.array([0.95]*1000)})
    expected_output = [{"x_min": i, "y_min": i+1, "x_max": i+2, "y_max": i+3, "class_id": "", "confidence": "", "tracker_id": "", "score": 0.95} for i in range(1000)]
    codeflash_output = CSVSink.parse_detection_data(detections)