Pixel statistics description#

The statistical description is assessed based on the image extracted from the interleaved calibration data. The pixel statistics description is done using the ctapipe-calculate-pixel-statistics tool. In order to produce test data for the camera calibration tools in the calibpipe package, the utility tool calibpipe-produce-camcalib-test-data can be used with the configuration files below.

The following configuration parameters are required:

  • allowed_tels: Optional list of allowed telescope IDs, others will be ignored.

  • input_column_name: Column name of the pixel-wise image data to calculate statistics.

  • output_table_name: Table name of the output statistics.

  • output_path: Output filename.

PixelStatisticsCalculator component is used to calculate the statistical description of the calibration events. It needs the following fields:

  • stats_aggregator_type: Name of the StatisticsAggregator subclass to be used.

  • outlier_detector_list: List of dicts containing the OutlierDetector and configuration to be used.

  • chunk_shift: Number of samples to shift the aggregation chunk for the calculation of the statistical values.

  • faulty_pixels_fraction: Minimum fraction of faulty camera pixels to identify regions of trouble.

Additionally, the user can provide the following fields for the StatisticsAggregator component:

  • PlainAggregator: Compute aggregated statistic values from a chunk of images using numpy functions.

  • SigmaClippingAggregator: Compute aggregated statistic values from a chunk of images using astropy’s sigma clipping functions.

Below are examples of configurations for assessing the statistical description of interleaved calibration events using the ctapipe-calculate-pixel-statistics tool.

Interleaved pedestal events - charge image

# Tool configuration for assess statistical description of pedestal events
PixelStatisticsCalculatorTool:
  input_column_name: image
# Pixel statistics configuration
PixelStatisticsCalculator:
  stats_aggregator_type:
    - ["type", "*", "SigmaClippingAggregator"]
  faulty_pixels_fraction: 0.1
  outlier_detector_list:
    -   name: StdOutlierDetector
        apply_to: median
        config:
            std_range_factors: [-10, 10]
    -   name: StdOutlierDetector
        apply_to: std
        config:
            std_range_factors: [-10, 10]
# Aggregation configuration
SigmaClippingAggregator:
  chunking_type: SizeChunking
  max_sigma: 4
  iterations: 5
# Chunking configuration
# Increase the chunk size to include more events per chunk for better statistics
SizeChunking:
  chunk_size: 50
  chunk_shift: 25

Interleaved flat-field events - charge image

# Tool configuration for assess statistical description of flatfield events
PixelStatisticsCalculatorTool:
  input_column_name: image
# Pixel statistics configuration
PixelStatisticsCalculator:
  stats_aggregator_type:
    - ["type", "*", "SigmaClippingAggregator"]
  faulty_pixels_fraction: 0.1
  outlier_detector_list:
    -   name: MedianOutlierDetector
        apply_to: median
        config:
            median_range_factors: [-0.9, 8]
    -   name: StdOutlierDetector
        apply_to: std
        config:
            std_range_factors: [-10, 10]
# Aggregation configuration
SigmaClippingAggregator:
  chunking_type: SizeChunking
  max_sigma: 4
  iterations: 5
# Chunking configuration
# Increase the chunk size to include more events per chunk for better statistics
SizeChunking:
  chunk_size: 50
  chunk_shift: 25

Interleaved flat-field events - peak arrival time

# Tool configuration for assess statistical description for the time correction
PixelStatisticsCalculatorTool:
  input_column_name: peak_time
# Pixel statistics configuration
PixelStatisticsCalculator:
  stats_aggregator_type:
    - ["type", "*", "PlainAggregator"]
  faulty_pixels_fraction: 0.1
  outlier_detector_list:
    -   name: RangeOutlierDetector
        apply_to: median
        config:
            validity_range: [2, 38]
# Aggregation configuration
PlainAggregator:
  chunking_type: SizeChunking
# Chunking configuration
# Increase the chunk size to include more events per chunk for better statistics
SizeChunking:
  chunk_size: 50
  chunk_shift: 25

This Page