Do not duplicate docs in config files

ctapipe/resources/train_disp_reconstructor.yaml

@@ -16,21 +16,23 @@ TrainDispReconstructor:
     rng_seed: 1337
 
   DispReconstructor:
-    prefix: disp  # Prefix of the output generated by the DispReconstructor (Default: "disp")
+    # prefix:  # Add a prefix of the output here, if you want to apply multiple
+               # DispReconstructors on the same file (e.g. for comparing different settings)
 
-    # All regression algorithms in scikit-learn are supported (https://scikit-learn.org/stable/modules/classes.html)
-    norm_cls: ExtraTreesRegressor  # scikit-learn algorithm to be used for the norm(disp) models
-    norm_config:                   # Settings passed to the scikit-learn algorithm
+    # All regression algorithms in scikit-learn are supported
+    # (https://scikit-learn.org/stable/modules/classes.html)
+    norm_cls: ExtraTreesRegressor
+    norm_config:
       n_estimators: 10
       max_depth: 10
       n_jobs: -1
 
-    log_target: True  # If true, norm(disp) models predict log(norm(disp)) and the output is
-                      # prefix_parameter = sign_prediction * exp(norm_prediction)
+    log_target: True
 
-    # All classification algorithms in scikit-learn are supported (https://scikit-learn.org/stable/modules/classes.html)
-    sign_cls: ExtraTreesClassifier  # scikit-learn algorithm to be used for the sign(disp) models
-    sign_config:                    # Settings passed to the scikit-learn algorithm
+    # All classification algorithms in scikit-learn are supported
+    # (https://scikit-learn.org/stable/modules/classes.html)
+    sign_cls: ExtraTreesClassifier
+    sign_config:
       n_estimators: 10
       max_depth: 10
       n_jobs: -1
@@ -45,7 +47,7 @@ TrainDispReconstructor:
         - ["not clipped", "leakage_intensity_width_2 < 0.5"]
 
     FeatureGenerator:  # On-the-fly generation of additional features
-      features:        # Supported functions: "sin", "cos", "tan", "arctan2", "log", "log10", "exp", "sqrt"
+      features:
       # - [<name>, <definition>]
         - ["area", "hillas_width * hillas_length"]
 
@@ -82,7 +84,7 @@ TrainDispReconstructor:
       - area  # Features generated by the FeatureGenerator still have to be listed here
               # for the models to use them.
 
-    stereo_combiner_cls: "StereoMeanCombiner"  # No other combiners are implemented atm.
-    StereoMeanCombiner:
-      weights: None     # Options: None, intensity, konrad
+    # stereo_combiner_cls: "StereoMeanCombiner"
+    # StereoMeanCombiner:  # Here you can set options for StereoMeanCombiner
+    #   weights:           # (e.g. which kind of weights to use).
 

ctapipe/resources/train_energy_regressor.yaml

@@ -16,14 +16,15 @@ TrainEnergyRegressor:
     rng_seed: 1337
 
   EnergyRegressor:
-    prefix: ExtraTreesRegressor  # Prefix of the output generated by the EnergyRegressor
-                                 # (Default: Name of the used scikit-learn algorithm).
-    log_target: True             # If true, models predict log(energy) and the output is
-                                 # prefix_energy = exp(model_prediction).
+    # prefix:  # Add a prefix of the output here, if you want to apply multiple
+               # EnergyRegressors on the same file (e.g. for comparing different settings)
 
-    # All regression algorithms in scikit-learn are supported (https://scikit-learn.org/stable/modules/classes.html)
-    model_cls: ExtraTreesRegressor  # scikit-learn algorithm to be used for the models
-    model_config:                   # Settings passed to the scikit-learn algorithm
+    log_target: True
+
+    # All regression algorithms in scikit-learn are supported
+    # (https://scikit-learn.org/stable/modules/classes.html)
+    model_cls: ExtraTreesRegressor
+    model_config:
       n_estimators: 10
       max_depth: 10
       n_jobs: -1
@@ -38,7 +39,7 @@ TrainEnergyRegressor:
         - ["not clipped", "leakage_intensity_width_2 < 0.5"]
 
     FeatureGenerator:  # On-the-fly generation of additional features
-      features:        # Supported functions: "sin", "cos", "tan", "arctan2", "log", "log10", "exp", "sqrt"
+      features:
       # - [<name>, <definition>]
         - ["area", "hillas_width * hillas_length"]
         - ["n_telescopes_triggered", "subarray.multiplicity(tels_with_trigger)"]
@@ -79,8 +80,7 @@ TrainEnergyRegressor:
       - n_telescopes_triggered             # still have to be listed here for the models
       - n_telescopes_hillas_reconstructor  # to use them.
 
-    stereo_combiner_cls: "StereoMeanCombiner"  # No other combiners are implemented atm.
-    StereoMeanCombiner:
-      weights: None     # Options: None, intensity, konrad
-      log_target: true  # If true, calculate exp(mean(log(values)))
+    # stereo_combiner_cls: "StereoMeanCombiner"
+    # StereoMeanCombiner:  # Here you can set options for StereoMeanCombiner
+    #   weights:           # (e.g. which kind of weights to use).
 

ctapipe/resources/train_particle_classifier.yaml

@@ -19,12 +19,13 @@ TrainParticleClassifier:
     rng_seed: 1337
 
   ParticleClassifier:
-    prefix: ExtraTreesClassifier  # Prefix of the output generated by the ParticleClassifier
-                                  # (Default: Name of the used scikit-learn algorithm).
+    # prefix:  # Add a prefix of the output here, if you want to apply multiple
+               # ParticleClassifiers on the same file (e.g. for comparing different settings)
 
-    # All classification algorithms in scikit-learn are supported (https://scikit-learn.org/stable/modules/classes.html)
-    model_cls: ExtraTreesClassifier  # scikit-learn algorithm to be used for the models
-    model_config:                    # Settings passed to the scikit-learn algorithm
+    # All classification algorithms in scikit-learn are supported 
+    # (https://scikit-learn.org/stable/modules/classes.html)
+    model_cls: ExtraTreesClassifier
+    model_config:
       n_estimators: 10
       max_depth: 10
       n_jobs: -1
@@ -39,7 +40,7 @@ TrainParticleClassifier:
         - ["not clipped", "leakage_intensity_width_2 < 0.5"]
 
     FeatureGenerator:  # On-the-fly generation of additional features
-      features:        # Supported functions: "sin", "cos", "tan", "arctan2", "log", "log10", "exp", "sqrt"
+      features:
       # - [<name>, <definition>]
         - ["area", "hillas_width * hillas_length"]
 
@@ -79,7 +80,7 @@ TrainParticleClassifier:
       - area  # Features generated by the FeatureGenerator still have to be listed here
               # for the models to use them.
 
-    stereo_combiner_cls: "StereoMeanCombiner"  # No other combiners are implemented atm.
-    StereoMeanCombiner:
-      weights: None     # Options: None, intensity, konrad
+    # stereo_combiner_cls: "StereoMeanCombiner"
+    # StereoMeanCombiner:  # Here you can set options for StereoMeanCombiner
+    #   weights:           # (e.g. which kind of weights to use).