pymrio parser walkthrough#

This walkthrough covers the mario.parse_from_pymrio(...) bridge from one in-memory pymrio.IOSystem to a mario.Database.

Main entry point#

  • mario.parse_from_pymrio(...)

  • upstream object: pymrio.IOSystem

Expected input and assignment patterns#

This parser expects one already-loaded pymrio.IOSystem, not a filesystem path.

Key arguments:

  • io: the pymrio.IOSystem to convert;

  • value_added: dictionary or "all" selector for the factor side;

  • satellite_account: dictionary or "all" selector for the satellite side;

  • include_meta: when True, copy the upstream metadata trail into the MARIO notes.

Both value_added and satellite_account can be one explicit dictionary or the shorthand "all". Dictionary values can also be "all" or one slicer for part of an extension. With the "all" / "all" shorthand, MARIO looks for one factor-like extension such as factor_inputs, factor_of_production, value_added, or primary_inputs, and assigns the remaining extensions to satellites.

The conversion result is an IOT, not a SUT.

[1]:

import mario
import pymrio

Parse one local EXIOBASE bundle with pymrio#

The path below is only an example. Replace it with your local EXIOBASE directory.

[2]:

io = pymrio.parse_exiobase3(
    path="/path/to/IOT_2013_pxp.zip",
)

Inspect available pymrio Extensions#

This helps you decide whether to pass explicit dictionaries or to rely on the "all" shorthand.

[3]:

[
    name
    for name in dir(io)
    if isinstance(getattr(io, name), pymrio.Extension)
]

[3]:

['air_emissions',
 'energy',
 'factor_inputs',
 'labour',
 'land',
 'material',
 'nutrients',
 'water']

Use the full shorthand#

When one unique factor-like Extension exists, you can use value_added="all" and satellite_account="all". This is practical for EXIOBASE parsed through pymrio, where factor_inputs is the natural value-added block.

[4]:

db = mario.parse_from_pymrio(
    io=io,
    value_added="all",
    satellite_account="all",
)

INFO Metadata: initialized.

INFO:mario.api.core_model:Metadata: initialized.

INFO Resolver: resolving v for baseline.

INFO:mario.compute.resolver:Resolver: resolving v for baseline.

INFO Resolver: trying v via formula build_iot_v_from_V_X.

INFO:mario.compute.resolver:Resolver: trying v via formula build_iot_v_from_V_X.

INFO Resolver: resolved X via formula build_iot_X_from_z_Y (compute_method=auto, runtime=solve).

INFO:mario.compute.resolver:Resolver: resolved X via formula build_iot_X_from_z_Y (compute_method=auto, runtime=solve).

INFO Resolver: resolved v via formula build_iot_v_from_V_X.

INFO:mario.compute.resolver:Resolver: resolved v via formula build_iot_v_from_V_X.

INFO Resolver: resolving e for baseline.

INFO:mario.compute.resolver:Resolver: resolving e for baseline.

INFO Resolver: trying e via formula build_iot_e_from_E_X.

INFO:mario.compute.resolver:Resolver: trying e via formula build_iot_e_from_E_X.

INFO Resolver: resolved e via formula build_iot_e_from_E_X.

INFO:mario.compute.resolver:Resolver: resolved e via formula build_iot_e_from_E_X.

INFO Resolver: resolving Z for baseline.

INFO:mario.compute.resolver:Resolver: resolving Z for baseline.

INFO Resolver: trying Z via formula build_iot_Z_from_z_X.

INFO:mario.compute.resolver:Resolver: trying Z via formula build_iot_Z_from_z_X.

INFO Resolver: resolved Z via formula build_iot_Z_from_z_X.

INFO:mario.compute.resolver:Resolver: resolved Z via formula build_iot_Z_from_z_X.

[5]:

db

[5]:

name = EXIO_IOT_2013_pxp
table = IOT
scenarios = ['baseline']
Factor of production = 9
Satellite account = 725
Consumption category = 7
Region = 49
Sector = 200

Reference notes and caveats#

Upstream documentation: pymrio documentation.

This is not a file parser. It converts one already-loaded pymrio.IOSystem, so any filesystem layout requirements belong to the upstream pymrio parser that created io.

Practical caveats:

  • every pymrio.Extension must be classified explicitly, either through one mapping or through the "all" shorthand;

  • the "all" / "all" shorthand works only when MARIO can infer one unique factor-like extension;

  • this is a conversion bridge, not a generic harmonizer between arbitrary pymrio objects and MARIO;

  • the resulting MARIO database is currently an IOT, not a SUT.

Download this notebook