Skip to content

Commit

Permalink
Merge branch 'main' of https://github.com/cafriedb/premise_validation
Browse files Browse the repository at this point in the history
  • Loading branch information
@cafriedb
cafriedb committed Aug 7, 2024
2 parents 3c75401 + 4b15b3f commit a50ad27
Show file tree
Hide file tree
Showing 7 changed files with 430 additions and 276 deletions.
22 changes: 10 additions & 12 deletions dopo/__init__.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,14 +5,12 @@
"__version__",
'generate_sets_from_filters',
'MethodFinder',
'find_and_create_method',
'get_all_methods',
'compare_activities_multiple_methods',
'lvl1_plot',
'lvl21_plot_stacked_absolute',
'lvl22_plot_input_comparison_with_method',
'lvl23_plot_input_comparison_plot_no_method',
'lvl3_plot_relative_changes'
'scores_across_activities',
'inputs_contributions',
'inputs_contribution',
'input_contribution_across_activities',
'activities_across_databases'
)

__version__ = "0.0.1"
Expand All @@ -22,9 +20,9 @@
from dopo.sector_score_dict import compare_activities_multiple_methods
from dopo.sector_score_dict import small_inputs_to_other_column
from dopo.plots import (
lvl1_plot,
lvl21_plot_stacked_absolute,
lvl22_plot_input_comparison_with_method,
lvl23_plot_input_comparison_plot_no_method,
lvl3_plot_relative_changes
scores_across_activities,
inputs_contributions,
inputs_contribution,
input_contribution_across_activities,
activities_across_databases
)
50 changes: 19 additions & 31 deletions dopo/activity_filter.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,11 @@
# Sector filter functions from premise
# ---------------------------------------------------


def _act_fltr(
database: list,
fltr = None,
mask = None,
fltr=None,
mask=None,
):
"""Filter `database` for activities_list matching field contents given by `fltr` excluding strings in `mask`.
`fltr`: string, list of strings or dictionary.
Expand Down Expand Up @@ -49,25 +50,22 @@ def _act_fltr(
if isinstance(value, list):
for val in value:
filters = [a for a in filters if val in a[field]]
#filters.extend([ws.either(*[ws.contains(field, v) for v in value])])

# filters.extend([ws.either(*[ws.contains(field, v) for v in value])])
else:
filters = [
a for a in filters if value in a[field]
]
filters = [a for a in filters if value in a[field]]

#filters.append(ws.contains(field, value))

# filters.append(ws.contains(field, value))

if mask:
for field, value in mask.items():
if isinstance(value, list):
for val in value:
filters = [f for f in filters if val not in f[field]]
#filters.extend([ws.exclude(ws.contains(field, v)) for v in value])
# filters.extend([ws.exclude(ws.contains(field, v)) for v in value])
else:
filters = [f for f in filters if value not in f[field]]
#filters.append(ws.exclude(ws.contains(field, value)))
# filters.append(ws.exclude(ws.contains(field, value)))

return filters

Expand All @@ -84,7 +82,7 @@ def generate_sets_from_filters(yaml_filepath, database=None) -> dict:
:rtype: dict
"""

filtr=_get_mapping(yaml_filepath, var='ecoinvent_aliases')
filtr = _get_mapping(yaml_filepath, var="ecoinvent_aliases")

names = []

Expand All @@ -98,35 +96,26 @@ def generate_sets_from_filters(yaml_filepath, database=None) -> dict:
else:
names.append(entry["fltr"])

#subset = list(
# subset = list(
# ws.get_many(
# database,
# ws.either(*[ws.contains("name", name) for name in names]),
# )
#)

subset=[
a for a in database if any(


x in a["name"] for x in names
)
]
# )

subset = [a for a in database if any(x in a["name"] for x in names)]

techs = {
tech: _act_fltr(subset, fltr.get("fltr"), fltr.get("mask"))
for tech, fltr in filtr.items()
}

mapping = {
tech: {act for act in actlst} for tech, actlst in techs.items()
}

mapping = {tech: {act for act in actlst} for tech, actlst in techs.items()}

return mapping

def _get_mapping(filepath, var):

def _get_mapping(filepath, var):
"""
Loa a YAML file and return a dictionary given a variable.
:param filepath: YAML file path
Expand All @@ -142,10 +131,9 @@ def _get_mapping(filepath, var):
for key, val in techs.items():
if var in val:
mapping[key] = val[var]

return mapping

return mapping

# Example on how to call the functions to create a set of filtered activities_list
#set_from_fltrs = generate_sets_from_filters(yaml_filepath, database=ei39SSP)

# Example on how to call the functions to create a set of filtered activities_list
# set_from_fltrs = generate_sets_from_filters(yaml_filepath, database=ei39SSP)
104 changes: 60 additions & 44 deletions dopo/cpc_inputs.py
View file
Original file line number Diff line number Diff line change
@@ -1,104 +1,120 @@

# Imports
# --------

#brightway
# brightway
import brightway2 as bw
import bw2analyzer as ba

#common
# common
import pandas as pd
import numpy as np

#to be completed
# to be completed
import ast

# Function to generate dataframes containing inputs in cpc format not characterized from an activity list
# Function to generate dataframes containing inputs in cpc format not characterized from an activity list
# Level 2.3 plot dependency
# ------------------------------------------------------------------------------------------------------------------------------------

def _get_cpc_inputs_of_activities(activities_list, input_type='list'):

'''
def _get_cpc_inputs_of_activities(activities_list, input_type="list"):
"""
for param description see function lvl23_plot_input_comparison_plot_no_method
NOTE: could adapt this function to get the outputs, or create another one. At the moment only inputs are considered.
'''
"""

def _activity_list_inputs_cpc(activities_list, input_type):
all_inputs = []
if input_type == 'list':

if input_type == "list":
activity_iterator = activities_list
elif input_type == 'dict':
elif input_type == "dict":
activity_iterator = activities_list.values()
else:
raise ValueError("input_type must be either 'list' or 'dict'")

for activity in activity_iterator:
inputs_keys = pd.Series({bw.get_activity(exc.input).key: exc.amount for exc in activity.technosphere()},
name=activity['name'] + ', ' + activity['location'])

inputs_keys = pd.Series(
{
bw.get_activity(exc.input).key: exc.amount
for exc in activity.technosphere()
},
name=activity["name"] + ", " + activity["location"],
)

# Adjust the way the key is presented
inputs_keys = inputs_keys.reset_index()
inputs_keys['full_key'] = inputs_keys.apply(lambda row: f"('{row['level_0']}', '{row['level_1']}')", axis=1)
inputs_keys = inputs_keys.drop(['level_0', 'level_1'], axis=1).set_index('full_key')

inputs_keys["full_key"] = inputs_keys.apply(
lambda row: f"('{row['level_0']}', '{row['level_1']}')", axis=1
)
inputs_keys = inputs_keys.drop(["level_0", "level_1"], axis=1).set_index(
"full_key"
)

# Add empty cpc column and activity information
inputs_keys.insert(0, 'identifier', activity['name'] + ', ' + activity['location'])
inputs_keys.insert(1, 'activity', activity['name'])
inputs_keys.insert(2, 'location', activity['location'])
inputs_keys.insert(3, 'unit', activity['unit'])
inputs_keys.insert(4, 'cpc', None)

inputs_keys.insert(
0, "identifier", activity["name"] + ", " + activity["location"]
)
inputs_keys.insert(1, "activity", activity["name"])
inputs_keys.insert(2, "location", activity["location"])
inputs_keys.insert(3, "unit", activity["unit"])
inputs_keys.insert(4, "cpc", None)

all_inputs.append(inputs_keys)

# Combine all inputs into a single DataFrame
combined_inputs = pd.concat(all_inputs, axis=0)

return combined_inputs

def _update_cpc_information(combined_inputs):
for index, row in combined_inputs.iterrows():
# Transform each key to tuple
tuple_key = ast.literal_eval(index)

# Get input activity for the key
input_activity = bw.get_activity(tuple_key)

# Get cpc name for activity
cpc_name = ba.comparisons.get_cpc(input_activity)

# Store cpc_name in the 'cpc' column of the combined_inputs dataframe
combined_inputs.at[index, 'cpc'] = cpc_name
combined_inputs.at[index, "cpc"] = cpc_name

return combined_inputs

def _transform_dataframe(combined_inputs):
# Set 'identifier' as the new index and drop the 'full_key' index
combined_inputs = combined_inputs.reset_index().set_index('identifier').drop('full_key', axis=1)

combined_inputs = (
combined_inputs.reset_index()
.set_index("identifier")
.drop("full_key", axis=1)
)

# Determine the index of the 'unit' column
unit_index = combined_inputs.columns.get_loc('unit')
unit_index = combined_inputs.columns.get_loc("unit")

# Split the dataframe into two parts
combined_inputs_identifier = combined_inputs.iloc[:, :unit_index+1]
combined_inputs_cpc = combined_inputs.iloc[:, unit_index+1:]
#set index of to 'cpc' in combined_input_cpc
combined_inputs_cpc = combined_inputs_cpc.set_index('cpc')
combined_inputs_identifier = combined_inputs.iloc[:, : unit_index + 1]
combined_inputs_cpc = combined_inputs.iloc[:, unit_index + 1 :]
# set index of to 'cpc' in combined_input_cpc
combined_inputs_cpc = combined_inputs_cpc.set_index("cpc")

# Combine rows with the same index value in combined_inputs_cpc
combined_inputs_cpc = combined_inputs_cpc.groupby(level=0).agg(lambda x: np.sum(x) if x.dtype.kind in 'biufc' else x.iloc[0])
combined_inputs_cpc = combined_inputs_cpc.groupby(level=0).agg(
lambda x: np.sum(x) if x.dtype.kind in "biufc" else x.iloc[0]
)
# Transpose combined_inputs_cpc
combined_inputs_cpc_trans = combined_inputs_cpc.T

# Merge combined_inputs_identifier and combined_inputs_cpc_trans
result = combined_inputs_identifier.join(combined_inputs_cpc_trans)
result = result.drop_duplicates()

# Sort dataframe by activity and location aplphabetically and reset the index
result = result.sort_values(by=['activity', 'location'])
result = result.sort_values(by=["activity", "location"])
result = result.reset_index(drop=True)
return result

Expand Down
Loading

0 comments on commit a50ad27

Please sign in to comment.