Quick Start

This minimal example walks you through performing a basic LCIA using edges.

import bw2data
from edges import EdgeLCIA, get_available_methods

bw2data.projects.set_current("some project")
act = bw2data.Database("some db").random()

# get available method
get_available_methods()

# Load a built-in method
lcia = EdgeLCIA(
    demand={act: 1},
    method=("AWARE 2.0", "Country", "all", "yearly")
)

# Step 1: Build the inventory
lcia.lci()

# Step 2.a: Match exchanges to characterization factors
lcia.map_exchanges()
# Step 2.b: since this is a regionalized method, a few more steps are required
lcia.map_aggregate_locations() # finds matches for aggregate regions ("RER", "US" etc.)
lcia.map_dynamic_locations() # finds matches for dynamic regions ("RoW", "RoW", etc.)
lcia.map_contained_locations() # finds matches for contained regions ("CA" for "CA-QC" if factor of "CA-QC" is not available)
lcia.map_remaining_locations_to_global() # applies global factors to remaining locations

# Step 3: Evaluate CFs (e.g., resolve symbolic expressions)
lcia.evaluate_cfs()

# Step 4: Compute the LCIA score
lcia.lcia()

# Step 5 (optional): Print a summary
print(lcia.statistics())

# Step 6 (optional but RECOMMENDED): Print a table with all exchanges characterized
# this allows you to check whether exchanges have been given the correct CFs
# include_unmatched=True allows you to see which exchanges were not matched (and if some should have been)
# split_aggregate_consumers=True expands weighted consumer fallback rows into countries
df = lcia.generate_cf_table(
    include_unmatched=False,
    split_aggregate_consumers=True,
)

For deterministic regionalized runs, split_aggregate_consumers=True replaces weighted fallback rows for consumer regions such as RER, GLO, RoW, and RoE with country-level rows in the exported table.

The same workflow also applies to mixed methods that combine biosphere-technosphere and technosphere-technosphere CF rows in one JSON file, such as the IBIF all pressures methods. In these runs, generate_cf_table() adds supplier matrix and direction columns so the two contribution families can be inspected separately.