cubie.systemmodels.SystemValues

Classes

SystemValues(values_dict, precision[, defaults])

A container for numerical values used to specify ODE systems.
class cubie.systemmodels.SystemValues.SystemValues(values_dict, precision, defaults=None, **kwargs)[source]

Bases: object

A container for numerical values used to specify ODE systems.

A container for numerical values such as initial state values, parameters, and observables (auxiliary variables). This object creates a corresponding array to feed to CUDA functions for compiling, and a dictionary of indices to look up that array.

Parameters:

  • values_array (np.ndarray) – Array containing the numerical values.

  • indices_dict (dict) – Dictionary mapping parameter names to array indices.

  • keys_by_index (dict) – Dictionary mapping array indices to parameter names.

  • values_dict (dict) – Dictionary containing parameter name-value pairs.

  • precision (np.dtype) – Data type for the values array.

Notes

If given a list of strings, it will create a dictionary with those strings as keys and 0.0 as the default value.

You can index into this object like a dictionary or an array, i.e. values[‘key’] or values[index or slice].

__init__(values_dict, precision, defaults=None, **kwargs)[source]

Initialize the system parameters.

Initialize the system parameters with default values, user-specified values from a dictionary, then any keyword arguments. Sets up an array of values and a dictionary mapping parameter names to indices.

Parameters:

  • values_dict (dict or list of str) – Full dictionary of parameter values, or dictionary of a subset of parameter values for use with a second dictionary of default values. This argument can also be a list of strings, in which case it will create a dictionary with those strings as keys.

  • precision (numpy.dtype) – Data type for the values array (e.g., np.float32, np.float64).

  • defaults (dict, optional) – Dictionary of default parameter values, if you’re only updating a subset.

  • **kwargs (dict) – Additional parameter values that override both defaults and values_dict.

Notes

If the same value occurs in the dict and keyword args, the kwargs one will win.

get_index_of_key(parameter_key, silent=False)[source]

Retrieve the index of a given key in the values_array.

Parameters:

  • parameter_key (str) – The parameter key to look up.

  • silent (bool, optional) – If True, suppresses KeyError if key is not found, by default False.

Returns:

The index of the parameter in the values_array.

Return type:

int

Raises:

  • KeyError – If parameter_key is not found and silent is False.

  • TypeError – If parameter_key is not a string.

get_indices(keys_or_indices, silent=False)[source]

Convert parameter identifiers to array indices.

Convert from the many possible forms that a user might specify parameters (str, int, list of either, array of indices) into an array of indices that can be passed to the CUDA functions.

Parameters:

  • keys_or_indices (str, int, slice, list, or numpy.ndarray) – Parameter identifiers that can be: - Single string parameter name - Single integer index - Slice object - List of string parameter names - List of integer indices - Numpy array of indices

  • silent (bool, optional) – Flag determining whether to raise KeyError if a key is not found, by default False.

Returns:

Array of parameter indices as np.int16 values.

Return type:

numpy.ndarray

Raises:

  • KeyError – If a string key is not found and silent is False.

  • IndexError – If an index is out of bounds.

  • TypeError – If the input type is not supported or mixed types are provided.

get_labels(indices)[source]

Get parameter labels corresponding to indices.

Parameters:

indices (list or numpy.ndarray) – Array or list of indices.

Returns:

List of labels corresponding to the provided indices.

Return type:

list of str

Raises:

TypeError – If indices is not a list or numpy array.

get_values(keys_or_indices)[source]

Retrieve parameter values.

Retrieve the value(s) of the parameter(s) from the values_dict. Accepts the same range of input types as get_indices.

Parameters:

keys_or_indices (str, int, list, or numpy.ndarray) – Parameter identifiers that can be: - Single string parameter name - Single integer index - List of string parameter names - List of integer indices - Numpy array of indices

Returns:

The parameter value(s) requested.

Return type:

float or numpy.ndarray

Raises:

  • KeyError – If a string key is not found in the parameters dictionary.

  • IndexError – If an integer index is out of bounds.

  • TypeError – If the input type is not supported.

indices_dict: dict
keys_by_index: dict
property names

Get parameter names.

Returns:

List of parameter names.

Return type:

list of str

precision: dtype
set_values(keys, values)[source]

Set parameter values.

Parameters:
Raises:

ValueError – If the number of keys does not match the number of values.

update_from_dict(values_dict, silent=False, **kwargs)[source]

Update dictionary and values_array with new values.

Updates both the values_dict and the values_array.

Parameters:

  • values_dict (dict) – Key-value pairs to update in the values_dict.

  • silent (bool, optional) – If True, suppresses KeyError if a key is not found in the parameters dictionary, by default False.

  • **kwargs (dict) – Additional key-value pairs to update.

Returns:

A set of keys that were successfully updated.

Return type:

set of str

Raises:

  • KeyError – If the key is not found in the parameters dictionary and silent is False.

  • TypeError – If any value in the values_dict cannot be cast to the specified precision.

update_param_array_and_indices()[source]

Extract values and create array and indices mapping.

Extract all values in self.values_dict and save to a numpy array with the specified precision. Also create a dict keyed by the values_dict key whose value is the index of the parameter in the created array.

Notes

The indices_dict dict will always be in the same order as the values_array, as long as the keys are extracted in the same order (insertion order is preserved in Python 3.7+).

values_array: ndarray
values_dict: dict