allow YAML and TOML config files and multiple configs

ctapipe/core/tool.py

@@ -1,21 +1,29 @@
 """Classes to handle configurable command-line user interfaces."""
 import logging
 import logging.config
-import textwrap
-from abc import abstractmethod
-import pathlib
 import os
+import pathlib
 import re
+import textwrap
+from abc import abstractmethod
+from typing import Union
+import yaml
 
-from traitlets import default
-from traitlets.config import Application, Configurable
+try:
+    import tomli as toml
+
+    HAS_TOML = True
+except ImportError:
+    HAS_TOML = False
+
+from traitlets import default, List
+from traitlets.config import Application, Config, Configurable
 
 from .. import __version__ as version
-from .traits import Path, Enum, Bool, Dict
 from . import Provenance
 from .component import Component
-from .logging import create_logging_config, ColoredFormatter, DEFAULT_LOGGING
-
+from .logging import DEFAULT_LOGGING, ColoredFormatter, create_logging_config
+from .traits import Bool, Dict, Enum, Path
 
 __all__ = ["Tool", "ToolConfigurationError"]
 
@@ -119,16 +127,21 @@ def main():
 
     """
 
-    config_file = Path(
-        exists=True,
-        directory_ok=False,
-        allow_none=True,
-        default_value=None,
-        help=(
-            "name of a configuration file with "
-            "parameters to load in addition to "
-            "command-line parameters"
-        ),
+    config_files = List(
+        trait=Path(
+            exists=True,
+            directory_ok=False,
+            allow_none=True,
+            default_value=None,
+            help=(
+                "List of configuration files with parameters to load "
+                "in addition to command-line parameters. "
+                "The order listed is the order of precendence (later config parameters "
+                "overwrite earlier ones), however parameters specified on the "
+                "command line always have the highest precendence. "
+                "Config files may be in JSON, YAML, TOML, or Python format"
+            ),
+        )
     ).tag(config=True)
 
     log_config = Dict(default_value=DEFAULT_LOGGING).tag(config=True)
@@ -158,7 +171,7 @@ def __init__(self, **kwargs):
         # make sure there are some default aliases in all Tools:
         super().__init__(**kwargs)
         aliases = {
-            ("c", "config"): "Tool.config_file",
+            ("c", "config"): "Tool.config_files",
             "log-level": "Tool.log_level",
             ("l", "log-file"): "Tool.log_file",
             "log-file-level": "Tool.log_file_level",
@@ -184,23 +197,54 @@ def __init__(self, **kwargs):
         self.update_logging_config()
 
     def initialize(self, argv=None):
-        """ handle config and any other low-level setup """
+        """handle config and any other low-level setup"""
         self.parse_command_line(argv)
         self.update_logging_config()
 
-        if self.config_file is not None:
-            self.log.debug(f"Loading config from '{self.config_file}'")
+        if self.config_files is not None:
+            self.log.info("Loading config from '%s'", self.config_files)
             try:
-                self.load_config_file(self.config_file)
+                for config_file in self.config_files:
+                    self.load_config_file(config_file)
             except Exception as err:
-                raise ToolConfigurationError(f"Couldn't read config file: {err}")
+                raise ToolConfigurationError(
+                    f"Couldn't read config file: {err} ({type(err)})"
+                ) from err
 
         # ensure command-line takes precedence over config file options:
         self.update_config(self.cli_config)
         self.update_logging_config()
 
         self.log.info(f"ctapipe version {self.version_string}")
 
+    def load_config_file(self, path: Union[str, pathlib.Path]) -> None:
+        """
+        Load a configuration file in one of the supported formats, and merge it with
+        the current config if it exists.
+
+        Parameters
+        ----------
+        path: Union[str, pathlib.Path]
+            config file to load. [yaml, toml, json, py] formats are supported
+        """
+
+        path = pathlib.Path(path)
+
+        if path.suffix in [".yaml", ".yml"]:
+            # do our own YAML loading
+            with open(path, "r") as infile:
+                config = Config(yaml.safe_load(infile))
+            self.update_config(config)
+        elif path.suffix == ".toml" and HAS_TOML:
+            with open(path, "rb") as infile:
+                config = Config(toml.load(infile))
+            self.update_config(config)
+        else:
+            # fall back to traitlets.config.Application's implementation
+            super().load_config_file(str(path))
+
+        Provenance().add_input_file(path, role="Tool Configuration")
+
     def update_logging_config(self):
         """Update the configuration of loggers."""
         cfg = create_logging_config(
@@ -330,11 +374,11 @@ def write_provenance(self):
 
     @property
     def version_string(self):
-        """ a formatted version string with version, release, and git hash"""
+        """a formatted version string with version, release, and git hash"""
         return f"{version}"
 
     def get_current_config(self):
-        """ return the current configuration as a dict (e.g. the values
+        """return the current configuration as a dict (e.g. the values
         of all traits, even if they were not set during configuration)
         """
         conf = {
@@ -350,7 +394,7 @@ def get_current_config(self):
         return conf
 
     def _repr_html_(self):
-        """ nice HTML rep, with blue for non-default values"""
+        """nice HTML rep, with blue for non-default values"""
         traits = self.traits()
         name = self.__class__.__name__
         lines = [
@@ -475,7 +519,6 @@ def run_tool(tool: Tool, argv=None, cwd=None):
         # switch to cwd for running and back after
         os.chdir(cwd)
         tool.run(argv or [])
-        return 0
     except SystemExit as e:
         return e.code
     finally:

ctapipe/resources/base_config.yaml

@@ -0,0 +1,111 @@
+# ==========================================================================
+# ctapipe-process configuration file.
+# version: VERSION
+#
+# This configuration contains a subset of options needed for a basic analysis.
+# Not all possible options are shown. To get a complete list, run:
+#
+#     `ctapipe-process --help-all`
+#
+# ==========================================================================
+
+DataWriter:
+  Contact:
+    # please fill in your contact information here. It will be stored in the
+    # output files as provenance information
+    name: YOUR-NAME-HERE
+    email: [email protected]
+    organization: YOUR-ORGANIZATION
+
+  # options that control what is stored in the output file by default here we
+  # write nothing (can be overridden on the command-line or in subsequent config
+  # files)
+  overwrite: false # do not overwrite existing files
+  write_images: false # store DL1 images
+  write_parameters: false # store DL1 parameters
+  write_stereo_shower: false # store DL2 stereo geometry
+  write_raw_waveforms: false # write R0 waveforms
+  write_waveforms: false # write R1 waveforms
+
+# The CameraCalibrator takes data from R1 or DL0 to DL1a level, applying finer
+# calibration and turning waveforms into images. It is run only if DL1a images
+# do not already exist in the input file.
+CameraCalibrator:
+  # Choose an extractor type from the following possibilities:
+  #'FullWaveformSum', 'FixedWindowSum', 'GlobalPeakWindowSum',
+  #'LocalPeakWindowSum', 'SlidingWindowMaxSum', 'NeighborPeakWindowSum',
+  #'TwoPassWindowSum', 'BaselineSubtractedNeighborPeakWindowSum'
+  #
+  # Note this is a telescope-wise parameter, so can be specified per telescope
+  # if necessary (see below for an example)
+  image_extractor_type: NeighborPeakWindowSum
+
+# The ImageProcessor performs the DL1a-> DL1b (image parameters) transition. It
+# is run only if the parameters `DataWriter.write_image_parameters=True` and the
+# parameters don't already exist in the input file (or if the user forces them
+# to be re-computed using DataWriter.recompute_dl1=True)
+ImageProcessor:
+  # The image cleaner selects pixels which have signal in them and rejects those
+  # without. Options are: 'TailcutsImageCleaner', 'MARSImageCleaner',
+  # 'FACTImageCleaner'
+  image_cleaner_type: TailcutsImageCleaner
+
+  # make sure you include a configuration for the image cleaner you selected
+  # above here. The section named by the image_cleaner_type will be used to
+  # configure it.
+  TailcutsImageCleaner:
+    # the thresholds for this image cleaner must be optimized for the data set
+    # you are analyzing. The defaults may not be correct and should be optimized
+    # for the given use case.
+    #
+    # These are telescope-wise parameters, where the options are patterns
+    # specified in a list in order of precedence, with later options overwriting
+    # earlier ones. Each pattern is a triplet of [scope, key, value], where the
+    # scope can be "type" (matching to the telescope type name) or "id"
+    # (matching a specific telescope ID number). In the case of "type", the key
+    # should be either a telescope type string, or part of one with "*" used as
+    # a wildcard match (e.g. "LST*" would match all LST telescopes). You can
+    # specify a universal default value using "*" as a key. Note that specifying
+    # a single value is equivalent to specifying a one-item list containing
+    # [type, '*', value] .
+    picture_threshold_pe: # top-level threshold in photoelectrons
+      - [type, "LST*", 6.0]
+      - [type, "MST*NectarCam", 8.0]
+      - [type, "MST*FlashCam", 10000] # disabled for now
+      - [type, "SST_ASTRI_CHEC", 4.0]
+    boundary_threshold_pe: # second-level threshold in photoelectrons
+      - [type, "LST*", 3.0]
+      - [type, "MST*NectarCam", 4.0]
+      - [type, "MST*FlashCam", 10000] # disabled for now
+      - [type, "SST_ASTRI_CHEC", 2.0]
+    keep_isolated_pixels: False # If False, pixels with  < min_picture_neighbors are removed.
+    min_picture_neighbors: 2 # Minimum number of neighbors above threshold to consider
+
+  # Choose which images should be parameterized:
+  ImageQualityQuery:
+    # quality critieria should be a list of pairs of [name,
+    # filter_function_string] The filter function should take a single value
+    # which is the image itself, a 1D np.array.
+    quality_criteria:
+      - ["enough_pixels", "lambda im: np.count_nonzero(im) > 2"]
+      - ["enough_charge", "lambda im: im.sum() > 50"]
+
+# The ShowerProcessor performs the DL1 to DL2a (reconstructed shower geometry)
+# transition. It is run only if the parameter DataWriter.write_stereo_shower=True.
+ShowerProcessor:
+  # choose between HillasReconstructor and HillasIntersection (two
+  # implementations of the standard stereo line-intersection method)
+  reconstructor_type: HillasReconstructor
+
+  HillasReconstructor:
+    # Choose which telescope events should be included in the reconstruction.
+    StereoQualityQuery:
+      # the quality criteria here should again be a list of [name,
+      # filter_function_string] pairs, with filter functions that take the set of
+      # image parameters, `p` (a `ctapipe.containers.ImageParametersContainer`), as
+      # an argument.
+      quality_criteria:
+        - [enough intensity, "lambda p: p.hillas.intensity > 50"]
+        - [Positive width, "lambda p: p.hillas.width.value > 0"]
+        - [enough pixels, "lambda p: p.morphology.num_pixels > 3"]
+        - [not clipped, "lambda p: p.leakage.intensity_width_2 < 0.8"]

ctapipe/resources/stage1_config.toml

@@ -0,0 +1,55 @@
+[DataWriter.Contact]
+# please fill in your contact information here. It will be stored in the
+# output files as provenance information
+name = "YOUR-NAME-HERE"
+email = "[email protected]"
+organization = "YOUR-ORGANIZATION"
+
+[DataWriter]
+# options that control what is stored in the output file
+overwrite = false
+write_images = true
+write_parameters = true
+write_stereo_shower = false
+write_mono_shower = false
+transform_image = true
+transform_peak_time = true
+
+[CameraCalibrator]
+image_extractor_type = "NeighborPeakWindowSum"
+
+[ImageProcessor]
+image_cleaner_type = "TailcutsImageCleaner"
+
+
+
+[ImageProcessor.TailcutsImageCleaner]
+picture_threshold_pe = [
+    [ "type", "*", 10.0,],
+    [ "type", "LST_LST_LSTCam", 5.0,],
+    [ "type", "MST_MST_NectarCam", 5.0,],
+    [ "type", "SST_ASTRI_CHEC", 3.0,],
+]
+boundary_threshold_pe = [
+    [ "type", "*", 5.0,],
+    [ "type", "LST_LST_LSTCam", 2.5,],
+    [ "type", "MST_MST_NectarCam", 2.5,],
+    [ "type", "SST_ASTRI_CHEC", 1.5,],
+]
+min_picture_neighbors = [ [ "type", "*", 2,],]
+
+[ImageProcessor.ImageQualityQuery]
+# These specify which images should be parameterized:
+quality_criteria = [
+    [ "enough_pixels", "lambda im: np.count_nonzero(im) > 2",],
+    [ "enough_charge", "lambda im: im.sum() > 50",]
+]
+
+[ShowerProcessor.ShowerQualityQuery]
+# These specify criteria for telescopes that should be included in stereo
+# reconstruction:
+quality_criteria = [
+    [ "enough intensity", "lambda p: p.hillas.intensity > 50",],
+    [ "Positive width", "lambda p: p.hillas.width.value > 0",],
+    [ "enough pixels", "lambda p: p.morphology.num_pixels > 3",],
+]

ctapipe/resources/stage1_config.yaml

@@ -0,0 +1,15 @@
+# ======================================================================
+# ctapipe-process configuration file.
+# version: VERSION
+#
+# Perform Stage 1 Processing
+#
+# This configuration enables options needed for Rx to DL1b (image parameters)
+# ======================================================================
+#
+#  Make sure you first include `--config base_config.yaml` before including this file
+
+DataWriter:
+  write_images: true
+  write_parameters: true
+  write_stereo_shower: false

ctapipe/resources/stage2_config.yaml

@@ -0,0 +1,16 @@
+# ======================================================================
+# ctapipe-process configuration file.
+# version: VERSION
+#
+# Perform Stage 2 Processing
+#
+# This configuration enables options needed for Rx to DL2a (shower
+# reconstruction) But not including any image parameters
+# ======================================================================
+#
+#  Make sure you first include `--config base_config.yaml` before including this file
+
+DataWriter:
+  write_images: false
+  write_parameters: false
+  write_stereo_shower: true

ctapipe/resources/training_config.yaml

@@ -0,0 +1,17 @@
+# ======================================================================
+# ctapipe-process configuration file.
+# version: VERSION
+#
+# Perform Stage 1 and Stage 2a processing to generate features needed for
+# training machine learning algorithms.
+#
+#  Note that here write_images is disabled by default (to avoid very large
+# files), however for training deep learning algorithms, it should be turned on.
+# ======================================================================
+#
+#  Make sure you first include `--config base_config.yaml` before including this file
+
+DataWriter:
+  write_images: false
+  write_parameters: true
+  write_stereo_shower: true