-
Notifications
You must be signed in to change notification settings - Fork 4
Track filtering
guijacquemet edited this page May 22, 2024
·
2 revisions
This section provides an interactive way to refine your tracking data. Here's what it's designed to achieve:
-Filter Tracks:
- Define a range for track lengths, mean speed, max speed, min speed, and total distance. This helps in filtering out tracks that might be artifacts or noise in your data.
-Smooth Tracks:
- Apply a moving average technique to the positional data in your tracks. By adjusting the
Smoothing Neighbors
slider, you can control the degree of smoothing. The smoothing of tracks is performed using a moving average technique, which averages the position data over a specified number of neighboring points centered around each data point. This reduces jitter and minor positional fluctuations in the data. For points at the edges where a full window of neighbors isn't available, the original values are used to ensure no data is lost.
- Track Duration Slider: Adjust to filter out tracks shorter than the specified duration.
- Mean Speed Slider: Adjust to filter out tracks with a mean speed less than the specified value.
- Max Speed Slider: Adjust to filter out tracks with a speed exceeding the specified value.
- Min Speed Slider: Adjust to filter out tracks with a speed below the specified value.
- Total Distance Slider: Adjust to filter out tracks that have traveled less than the specified distance.
- Smoothing Neighbors Slider: Set the number of neighboring points used for smoothing the tracks. Higher values result in smoother tracks by averaging over more points.
- Adjust the sliders to your desired filtering and smoothing criteria.
- Apply Filters: After adjusting the sliders, click the "Apply Filters" button. This will process the data based on your settings and prepare it for downstream analyses.
The filtering and smoothing parameters are saved in a CSV file in the results folder, allowing for reproducibility and easy reference for future analyses.
After refining your data, it's pivotal to decide which dataset will serve as the foundation for subsequent analysis:
- Visual Comparison: Contrast raw and filtered tracks side-by-side to make an informed decision.
This section allows you to select whether to use your filtered and smoothed dataset or the raw dataset for downstream analysis. If you choose the filtered dataset, raw tracks will be discarded beyond this point. To recover them, you will need to reload your dataset.
- Data Selection: Use the radio buttons to select either the raw data or the smoothed and filtered data for further analysis.
- Confirmation: Click the "Select" button to confirm your choice.
---