oasislmf.pytools.fm.financial_structure

Functions

load_static(static_path)

Load the raw financial data from static_path as numpy ndarray

extract_financial_structure(allocation_rule, ...)

Build the in-memory financial structure arrays from the raw fm input files.

create_financial_structure(allocation_rule, static_path)

Compute the financial structure and save it as .npy files in static_path.

load_financial_structure(allocation_rule, static_path)

Module Contents

oasislmf.pytools.fm.financial_structure.load_static(static_path)[source]

Load the raw financial data from static_path as numpy ndarray first check if .bin file is present then try .cvs try loading profile_step before falling back to normal profile,

Parameters:

static_path (str) – path to the folder holding the static input files

Returns:

  • programme: link between nodes

  • policytc: info on layer

  • profile: policy profile can be profile_step or profile

  • stepped: True when fm_profile_step was loaded, None when falling back to fm_profile

  • xref: node to output_id

  • items: items (item_id and coverage_id mapping), empty when items and coverages disagree on the number of coverages

  • coverages: Tiv value for each coverage id, empty when items and coverages disagree on the number of coverages

Return type:

Tuple[np.ndarray, np.ndarray, np.ndarray, Optional[bool], np.ndarray, np.ndarray, np.ndarray]

Raises:

FileNotFoundError – if one of the static is missing

oasislmf.pytools.fm.financial_structure.extract_financial_structure(allocation_rule, fm_programme, fm_policytc, fm_profile, stepped, fm_xref, items, coverages)[source]

Build the in-memory financial structure arrays from the raw fm input files.

Parameters:

  • allocation_rule (int) – option to indicate how the losses are allocated to the output

  • fm_programme (np.ndarray[fm_programme_dtype]) – structure of the levels

  • fm_policytc (np.ndarray[fm_policytc_dtype]) – structure of the layers and policy_id to apply

  • fm_profile (np.ndarray) – definition of the policy_id, of fm_profile_dtype or fm_profile_step_dtype

  • stepped (Optional[bool]) – True when fm_profile holds step policies, None otherwise

  • fm_xref (np.ndarray[fm_xref_dtype]) – mapping between the output of the allocation and output item_id

  • items (np.ndarray[items_dtype]) – item_id and coverage_id mapping, empty when unavailable

  • coverages (np.ndarray[oasis_float]) – Tiv value for each coverage id, empty when unavailable

Returns:

  • compute_infos: array describing the steps of the computation to perform

  • nodes_array: array of the nodes of the financial structure

  • node_parents_array: array mapping each node to its parent nodes

  • node_profiles_array: array mapping each node to its profiles

  • output_array: array mapping each output to its node

  • fm_profile: the fm_profile array used by the computation

Return type:

Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray]

Raises:

KeyError – if a node at the output level has no output_id

oasislmf.pytools.fm.financial_structure.create_financial_structure(allocation_rule, static_path)[source]

Compute the financial structure and save it as .npy files in static_path.

The extracted structure (compute_info, nodes_array, node_parents_array, node_profiles_array, output_array and fm_profile) is written to static_path; nothing is returned.

Parameters:

  • allocation_rule (int) – back-allocation rule

  • static_path (str) – path to the static files

Raises:

ValueError – if allocation_rule is not one of the allowed allocation rules

oasislmf.pytools.fm.financial_structure.load_financial_structure(allocation_rule, static_path)[source]