Remove useless namespaces / change runcard names

doc/sphinx/source/n3fit/runcard_detailed.rst

@@ -74,17 +74,16 @@ It is possible to run a fit with no validation set by setting the fraction to ``
 Network Architecture
 --------------------
 There are different network architectures implemented in ``n3fit``.
-Which can be selected by changing the ``fitting:parameters::layer_type`` parameter in the runcard.
+Which can be selected by changing the ``parameters::layer_type`` parameter in the runcard.
 All layer types implement the ``nodes_per_layer``, ``activation_per_layer`` and ``initializer`` parameters.
 
 .. code-block:: yaml
 
-    fitting:
-        parameters:
-            nodes_per_layer: [5, 3, 8]
-            activation_per_layer: ['tanh', 'tanh', 'linear']
-            layer_type: 'dense_per_flavour'
-            initializer: 'glorot_normal'
+    parameters:
+        nodes_per_layer: [5, 3, 8]
+        activation_per_layer: ['tanh', 'tanh', 'linear']
+        layer_type: 'dense_per_flavour'
+        initializer: 'glorot_normal'
 
 - **One single network** (``layer_type: dense``):
 
@@ -117,12 +116,11 @@ of optimizer (and its corresponding options).
 
 .. code-block:: yaml
 
-    fitting:
-        parameters:
-            optimizer:
-              optimizer_name: 'Adadelta'
-              learning_rate: 1.0
-              clipnorm: 1.0
+    parameters:
+        optimizer:
+          optimizer_name: 'Adadelta'
+          learning_rate: 1.0
+          clipnorm: 1.0
 
 
 The full list of optimizers accepted by the ``n3fit`` and their arguments
@@ -136,13 +134,13 @@ Positivity
 ----------
 
 In ``n3fit`` the behavior of the positivity observables has changed with respect to ``nnfit``.
-In ``nnfit`` the loss due to the positivity observable was multiplied by a ``poslambda`` for each observable, defined in the runcard as:
+In ``nnfit`` the loss due to the positivity observable was multiplied by a ``maxlambda`` for each observable, defined in the runcard as:
 
 .. code-block:: yaml
 
     positivity:
       posdatasets:
-        - {dataset: POSF2U, poslambda: 1e6}
+        - {dataset: POSF2U, maxlambda: 1e6}
 
 
 This behavior was found to be very inefficient for gradient descent based strategies and was exchanged for a dynamical Lagrange multiplier.
@@ -152,19 +150,18 @@ the Neural Network as:
 
 .. code-block:: yaml
 
-    fitting:
-        parameters:
-            positivity:
-              threshold: 1e-6
-              multiplier: 1.05
-              initial: 14.5
+    parameters:
+        positivity:
+          threshold: 1e-6
+          multiplier: 1.05
+          initial: 14.5
 
 Note that by defining the positivity in this way all datasets will share the same Lagrange multiplier.
 
 It is also possible to not define the positivity hyperparameters (or define them only partially).
 In this case ``n3fit`` will set the initial Lagrange multiplier as ``initial`` (default: 1.0)
 while the ``multiplier`` will be such that after the last epoch the final Lagrange multiplier 
-equals the ``poslambda`` defined for the dataset.
+equals the ``maxlambda`` defined for the dataset.
 
 Finally we have the positivity threshold, which is set to ``1e-6`` by default.
 During the fit, the positivity loss will be compared to this value. If it is above it,
@@ -200,9 +197,8 @@ Threshold :math:`\chi2`
 
 .. code-block:: yaml
 
-    fitting:
-        parameters:
-            threshold_chi2: 4.0
+    parameters:
+        threshold_chi2: 4.0
 
 - ``threshold_chi2``: sets a maximum validation :math:`\chi2` for the stopping to activate. Avoids (too) early stopping.
 
@@ -220,10 +216,9 @@ In order to enable the TensorBoard callback in ``n3fit`` it is enough with addin
 
 .. code-block:: yaml
 
-    fitting:
-        tensorboard:
-            weight_freq: 100
-            profiling: True
+    tensorboard:
+        weight_freq: 100
+        profiling: True
 
 
 The ``weight_freq`` flag controls each how many epochs the weights of the NN are stored.
@@ -267,9 +262,8 @@ Threshold :math:`\chi2`
 
 .. code-block:: yaml
 
-    fitting:
-        parameters:
-            threshold_chi2: 4.0
+    parameters:
+        threshold_chi2: 4.0
 
 - ``threshold_chi2``: sets a maximum validation :math:`\chi2` for the stopping to activate. Avoids (too) early stopping.
 
@@ -279,9 +273,8 @@ Save and load weights of the model
 
 .. code-block:: yaml
 
-    fitting:
-        save: "weights.h5"
-        load: "weights.h5"
+    save: "weights.h5"
+    load: "weights.h5"
 
 - ``save``: saves the weights of the PDF model in the selected file in the replica folder.
 - ``load``: loads the weights of the PDF model from the selected file.

doc/sphinx/source/tutorials/pseudodata.md

@@ -2,7 +2,7 @@
 
 Suppose one has obtained a fit using the N3FIT framework and wants to do some analysis that requires
 knowing exactly the data that the neural networks saw during the fitting procedure. Thankfully, this
-information is reproducible due to the various seeds in the `fitting` namespace in the fit runcard.
+information is reproducible due to the various seeds in the fit runcard.
 
 The 3 seeds of interest are `trvlseed` which determines the training/validation splitting, `nnseed`
 which concerns the initialization of the neural netowrks themselves, and finally `mcseed` which is the

doc/sphinx/source/tutorials/run-fit.md

@@ -28,40 +28,34 @@ experimental data, the theory setup and the fitting setup.
 The `n3fit` code accepts the same YAML keys used by `nnfit` and extra keys
 required by the new techniques introduced in [Methodology](methodology).
 
-In particular we introduce new keys in the `fitting` section such as:
+In particular we introduce new keys in the fit runcard such as:
 - different seeds for the training and validation split (`trvlseed`), neural
 network intialization (`nnseed`) and the MC data replica generation (`mcseed`).
-- some debug flags to store or load the model in/from hd5 files (`save`,
-`savefile`, `load`, `loadfile`, `plot`)
+- some debug flags to store or load the model in/from hd5 files (`save`, `load`)
 - a new `parameters` dictionnary with the model specifications based on the
 arguments described in [Methodology](methodology).
 
 See, as an example, the following self explanatory runcard fragment:
 ```yaml
 # runcard example
 ...
-fitting:
-  trvlseed: 1
-  nnseed: 2
-  mcseed: 3
-  save: False
-  savefile: 'weights.hd5'
-  load: False
-  loadfile: 'weights.hd5'
-  plot: False
-
-  parameters: # This defines the parameter dictionary that is passed to the Model Trainer
-    nodes_per_layer: [15, 10, 8]
-    activation_per_layer: ['sigmoid', 'sigmoid', 'linear']
-    initializer: 'glorot_normal'
-    learning_rate: 0.01
-    optimizer: 'RMSprop'
-    epochs: 900
-    pos_multiplier: 1.05
-    pos_initial:  # believe the pos_lambda below
-    stopping_patience: 0.30 # percentage of the number of epochs
-    layer_type: 'dense'
-    dropout: 0.0
+trvlseed: 1
+nnseed: 2
+mcseed: 3
+genrep: true
+
+parameters: # This defines the parameter dictionary that is passed to the Model Trainer
+  nodes_per_layer: [15, 10, 8]
+  activation_per_layer: ['sigmoid', 'sigmoid', 'linear']
+  initializer: 'glorot_normal'
+  learning_rate: 0.01
+  optimizer: 'RMSprop'
+  epochs: 900
+  pos_multiplier: 1.05
+  pos_initial:  # believe the pos_lambda below
+  stopping_patience: 0.30 # percentage of the number of epochs
+  layer_type: 'dense'
+  dropout: 0.0
 ...
 ```
 

doc/validphys2/guide.md

@@ -2848,7 +2848,7 @@ resource we are providing does. The docstring will be seen in
 ```python
 def parse_posdataset(self, posset:dict, * ,theoryid):
     """An observable used as positivity constrain in the fit.
-    It is a mapping containing 'dataset' and 'poslambda'."""
+    It is a mapping containing 'dataset' and 'maxlambda'."""
     ...
 ```
 
@@ -2938,10 +2938,10 @@ errors (see the other examples in the class).
 The `PositivytySetSpec` could be defined roughly like:
 ```python
  class PositivitySetSpec():
-     def __init__(self, commondataspec, fkspec, poslambda, thspec):
+     def __init__(self, commondataspec, fkspec, maxlambda, thspec):
          self.commondataspec = commondataspec
          self.fkspec = fkspec
-         self.poslambda = poslambda
+         self.maxlambda = maxlambda
          self.thspec = thspec
 
      @property
@@ -2955,7 +2955,7 @@ The `PositivytySetSpec` could be defined roughly like:
      def load(self):
          cd = self.commondataspec.load()
          fk = self.fkspec.load()
-         return PositivitySet(cd, fk, self.poslambda)
+         return PositivitySet(cd, fk, self.maxlambda)
 ```
 Here `PositivitySet` is the `libnnpdf` object. It is generally better
 to pass around the spec objects because they are lighter and have more
@@ -2966,21 +2966,21 @@ With this, our parser method could look like this:
 
 def parse_posdataset(self, posset:dict, * ,theoryid):
     """An observable used as positivity constrain in the fit.
-    It is a mapping containing 'dataset' and 'poslambda'."""
+    It is a mapping containing 'dataset' and 'maxlambda'."""
     bad_msg = ("posset must be a mapping with a name ('dataset') and "
-               "a float multiplier(poslambda)")
+               "a float multiplier(maxlambda)")
 
     theoryno, theopath = theoryid
     try:
         name = posset['dataset']
-        poslambda = float(posset['poslambda'])
+        maxlambda = float(posset['maxlambda'])
     except KeyError as e:
         raise ConfigError(bad_msg, e.args[0], posset.keys()) from e
     except ValueError as e:
         raise ConfigError(bad_msg) from e
 
     try:
-        return self.loader.check_posset(theoryno, name, poslambda)
+        return self.loader.check_posset(theoryno, name, maxlambda)
     except FileNotFoundError as e:
         raise ConfigError(e) from e
 ```

n3fit/runcards/Basic_feature_scaling.yml

@@ -34,40 +34,37 @@ theory:
   theoryid: 53        # database id
 
 ############################################################
-fitting:
-  trvlseed: 1
-  nnseed: 2
-  mcseed: 3
-  epochs: 900
-  save: 'weights.h5'
-#   load: '/path/to/weights.h5/file'
-
+trvlseed: 1
+nnseed: 2
+mcseed: 3
+genrep   : True     # true = generate MC replicas, false = use real data
 
-  tensorboard:
-    weight_freq: 100
-    profiling: False
+save: 'weights.h5'
 
-  genrep   : True     # true = generate MC replicas, false = use real data
+tensorboard:
+  weight_freq: 100
+  profiling: False
 
-  parameters: # This defines the parameter dictionary that is passed to the Model Trainer
-    nodes_per_layer: [15, 10, 8]
-    activation_per_layer: ['sigmoid', 'sigmoid', 'linear']
-    initializer: 'glorot_normal'
-    optimizer:
-      optimizer_name: 'RMSprop'
-      learning_rate: 0.01
-      clipnorm: 1.0
-    epochs: 900
-    positivity:
-      multiplier: 1.05 # When any of the multiplier and/or the initial is not set
-      initial: # the poslambda will be used instead to compute these values per dataset
-      threshold: 1e-5
-    stopping_patience: 0.30 # percentage of the number of epochs
-    layer_type: 'dense'
-    dropout: 0.0
-    interpolation_points: 40
-    threshold_chi2: 5.0
+parameters: # This defines the parameter dictionary that is passed to the Model Trainer
+  nodes_per_layer: [15, 10, 8]
+  activation_per_layer: ['sigmoid', 'sigmoid', 'linear']
+  initializer: 'glorot_normal'
+  optimizer:
+    optimizer_name: 'RMSprop'
+    learning_rate: 0.01
+    clipnorm: 1.0
+  epochs: 900
+  positivity:
+    multiplier: 1.05 # When any of the multiplier and/or the initial is not set
+    initial: # the maxlambda will be used instead to compute these values per dataset
+    threshold: 1e-5
+  stopping_patience: 0.30 # percentage of the number of epochs
+  layer_type: 'dense'
+  dropout: 0.0
+  interpolation_points: 40
+  threshold_chi2: 5.0
 
+fitting:
   # NN23(QED) = sng=0,g=1,v=2,t3=3,ds=4,sp=5,sm=6,(pht=7)
   # EVOL(QED) = sng=0,g=1,v=2,v3=3,v8=4,t3=5,t8=6,(pht=7)
   # EVOLS(QED)= sng=0,g=1,v=2,v8=4,t3=4,t8=5,ds=6,(pht=7)
@@ -79,25 +76,25 @@ fitting:
       # mutsize: mutation size
       # mutprob: mutation probability
       # smallx, largex: preprocessing ranges
-      - { fl: sng, pos: False, mutsize: [15], mutprob: [0.05], smallx: [1.05,1.19], largex: [1.47,2.70], trainable: False }
-      - { fl: g,   pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.94,1.25], largex: [0.11,5.87], trainable: False }
-      - { fl: v,   pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.54,0.75], largex: [1.15,2.76], trainable: False }
-      - { fl: v3,  pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.21,0.57], largex: [1.35,3.08] }
-      - { fl: v8,  pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.52,0.76], largex: [0.77,3.56], trainable: True }
-      - { fl: t3,  pos: False, mutsize: [15], mutprob: [0.05], smallx: [-0.37,1.52], largex: [1.74,3.39] }
-      - { fl: t8,  pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.56,1.29], largex: [1.45,3.03] }
-      - { fl: cp,  pos: False, mutsize: [15], mutprob: [0.05], smallx: [0.12,1.19], largex: [1.83,6.70] }
+      - { fl: sng, pos: False, smallx: [1.05,1.19], largex: [1.47,2.70], trainable: False }
+      - { fl: g,   pos: False, smallx: [0.94,1.25], largex: [0.11,5.87], trainable: False }
+      - { fl: v,   pos: False, smallx: [0.54,0.75], largex: [1.15,2.76], trainable: False }
+      - { fl: v3,  pos: False, smallx: [0.21,0.57], largex: [1.35,3.08] }
+      - { fl: v8,  pos: False, smallx: [0.52,0.76], largex: [0.77,3.56], trainable: True }
+      - { fl: t3,  pos: False, smallx: [-0.37,1.52], largex: [1.74,3.39] }
+      - { fl: t8,  pos: False, smallx: [0.56,1.29], largex: [1.45,3.03] }
+      - { fl: cp,  pos: False, smallx: [0.12,1.19], largex: [1.83,6.70] }
 
 ############################################################
 positivity:
   posdatasets:
-    - { dataset: POSF2U,   poslambda: 1e6 }  # Positivity Lagrange Multiplier
-    - { dataset: POSFLL,   poslambda: 1e4 }
+    - { dataset: POSF2U,   maxlambda: 1e6 }  # Positivity Lagrange Multiplier
+    - { dataset: POSFLL,   maxlambda: 1e4 }
 
 ############################################################
 integrability:
   integdatasets:
-    - {dataset: INTEGXT3,   poslambda: 1e2}
+    - {dataset: INTEGXT3,   maxlambda: 1e2}
 
 ############################################################
 debug: True