pydrobert-param 0.0.0

Utilities for the python package 'param'

# pydrobert-param
Utilities for the python package [param](http://param.pyviz.org/)

## Serialization
While _param_ is a great tool for keeping track of parameters, currently
serialization is restricted to pickling and unpickling objects.
_pydrobert-param_ takes the stance that parameter (de)serialization has an
intuitive and obvious representation in most cases. The developer can modify
how _pydrobert-param_ (de)serializes `param.Parameterized` instances according
to her needs.

As an example, suppose we have parameterized classes and instances:
``` python
import param
class TrainingHyperparameters(param.Parameterized):
lr = param.Number(1e-5, doc='The learning rate')
max_epochs = param.Integer(10)
model_regex = param.String(
"model-{epoch:05d}.pkl",
doc='Regular expression for storing model weights after every epoch')

t_params = TrainingHyperparameters()

class ModelHyperparameters(param.Parameterized):
layers = param.ListSelector(
[], objects=['conv', 'fc', 'recurrent'],
doc='Sequence of layers by type, bottom-first')
activations = param.ObjectSelector('relu', objects=['tanh', 'relu'])

m_params = ModelHyperparameters()
m_params.layers = ['conv', 'conv', 'fc']

param_dict = {
'training': t_params,
'model': m_params,
}
```
We can serialize these easily into JSON or YAML:
``` python
from pydrobert.param import serialization as serial
serial.serialize_to_json('conf.json', param_dict)
# requires ruamel.yaml or pyyam
serial.serialize_to_yaml('conf.yaml', param_dict)
```
where we get
``` json
{
"training": {
"lr": 1e-05,
"max_epochs": 10,
"model_regex": "model-{epoch:05d}.pkl"
},
"model": {
"layers": [
"conv",
"conv",
"fc"
],
"activations": "relu"
}
}
```
and
``` yaml
training:
lr: 1e-05 # The learning rate
max_epochs: 10
model_regex: model-{epoch:05d}.pkl # Regular expression for storing model weights after every epoch
model:
layers: # Sequence of layers by type, bottom-first. Element choices: "conv", "fc", "recurrent"
- conv
- conv
- fc
activations: relu # Choices: "tanh", "relu"
```
respectively.

Deserialization proceeds similarly. Files can be used to populate parameters in
existing parameterized instances.
``` python
t_params.lr = 10000.
assert t_params.lr == 10000.
serial.deserialize_from_yaml('conf.yaml', param_dict)
assert t_params.lr == 1e-05
```
_pydrobert.param.argparse_ contains convenience methods for deserializing
config files right from the command line. Wow, neat-o!

Sometimes, the default (de)serialization routines are unsuited for the data.
For example, INI files do not have a standard format for lists of values. We
can design a custom serializer and deserializer for handling our `"layers"`
parameter:
``` python
class CommaSerializer(serial.DefaultListSelectorSerializer):
def help_string(self, name, parameterized):
choices_help_string = super(CommaSerializer, self).help_string(name, parameterized)
return 'Elements separated by commas. ' + choices_help_string

def serialize(self, name, parameterized):
val = super(CommaSerializer, self).serialize(name, parameterized)
return ','.join(str(x) for x in val)

class CommaDeserializer(serial.DefaultListSelectorDeserializer):
def deserialize(self, name, block, parameterized):
block = block.split(',')
super(CommaDeserializer, self).deserialize(name, block, parameterized)

serial.serialize_to_ini(
'conf.ini', param_dict,
serializer_name_dict={'model': {'layers': CommaSerializer()}},
)
serial.deserialize_from_ini(
'conf.ini', param_dict,
deserializer_name_dict={'model': {'layers': CommaDeserializer()}},
)
```
With _conf.ini_:
``` INI
# == Help ==
# [training]
# lr: The learning rate
# model_regex: Regular expression for storing model weights after every epoch

# [model]
# activations: Choices: "tanh", "relu"
# layers: Sequence of layers by type, bottom-first. Elements separated by commas. Element choices: "conv", "fc", "recurrent"


[training]
max_epochs = 10
lr = 1e-05
model_regex = model-{epoch:05d}.pkl

[model]
activations = relu
layers = conv,conv,fc
```