Outcomes

models.Outcomes(data, *, effect='qaly', comparator=None)

Probabilistic sensitivity analysis outcomes per intervention per iteration.

Parameters

Name	Type	Description	Default
data	pd.DataFrame	DataFrame indexed by a two-level MultiIndex named ("intervention", "iteration"), with a "cost" column and at least the primary effect column. Any additional numeric columns are carried along as disaggregated components.	required
effect	str	Name of the primary effect column (default "qaly").	'qaly'
comparator	str | None	Name of the reference intervention (the PICOTS comparator), or None if no intervention was flagged is_comparator=True. heormodel.cea.ce_plane and heormodel.report.tornado_data fall back to this, then to the first intervention, when their own comparator argument is omitted.	None

Example

import pandas as pd from heormodel.models import Outcomes tidy = pd.DataFrame({ … “intervention”: [“A”, “A”, “B”, “B”], … “iteration”: [0, 1, 0, 1], … “cost”: [100.0, 110.0, 200.0, 190.0], … “qaly”: [1.0, 1.1, 1.4, 1.3], … }) out = Outcomes.from_tidy(tidy) out.interventions [‘A’, ‘B’] out.n_iterations 2

Attributes

Name	Description
components	Names of disaggregated component columns beyond cost and primary effect.
interventions	Intervention names in first-appearance order.
iterations	The shared iteration index.
n_iterations	Number of iterations.

Methods

Name	Description
costs_wide	Costs as an (iterations x interventions) matrix.
effects_wide	An effect column as an (iterations x interventions) matrix.
from_tidy	Build from a tidy long table (the bring-your-own-outputs entry point).
from_wide	Build from two wide tables (iterations x interventions).
select	Subset to the given interventions, preserving the iteration index.
summary	Mean of every outcome column per intervention.

costs_wide

models.Outcomes.costs_wide()

Costs as an (iterations x interventions) matrix.

Example

import pandas as pd from heormodel.models import Outcomes c = pd.DataFrame({“A”: [1.0], “B”: [2.0]}) e = pd.DataFrame({“A”: [0.1], “B”: [0.2]}) Outcomes.from_wide(c, e).costs_wide().shape (1, 2)

effects_wide

models.Outcomes.effects_wide(column=None)

An effect column as an (iterations x interventions) matrix.

Parameters

Name	Type	Description	Default
column	str | None	Effect column name; defaults to the primary effect.	None

from_tidy

models.Outcomes.from_tidy(
    df,
    *,
    intervention='intervention',
    iteration='iteration',
    cost='cost',
    effect='qaly',
    comparator=None,
)

Build from a tidy long table (the bring-your-own-outputs entry point).

Parameters

Name	Type	Description	Default
df	pd.DataFrame	Long table with one row per (intervention, iteration).	required
intervention	str	Column in df holding the intervention label.	'intervention'
iteration	str	Column in df holding the iteration index.	'iteration'
cost	str	Column in df holding the cost per iteration.	'cost'
effect	str	Column in df holding the effect (QALYs by default).	'qaly'
comparator	str | None	Name of the reference intervention, or None.	None

Example

import pandas as pd from heormodel.models import Outcomes df = pd.DataFrame({“arm”: [“A”, “B”], “iter”: [0, 0], … “cost”: [1.0, 2.0], “qaly”: [0.5, 0.6]}) Outcomes.from_tidy(df, intervention=“arm”, iteration=“iter”).n_iterations 1

from_wide

models.Outcomes.from_wide(costs, effects, *, effect='qaly', comparator=None)

Build from two wide tables (iterations x interventions).

Parameters

Name	Type	Description	Default
costs	pd.DataFrame	Costs, one column per intervention, indexed by iteration.	required
effects	pd.DataFrame	Effects with identical shape/labels.	required
effect	str	Name to give the effect column.	'qaly'
comparator	str | None	Name of the reference intervention, or None.	None

Example

import pandas as pd from heormodel.models import Outcomes c = pd.DataFrame({“A”: [1.0, 2.0], “B”: [3.0, 4.0]}) e = pd.DataFrame({“A”: [0.1, 0.2], “B”: [0.3, 0.4]}) Outcomes.from_wide(c, e).interventions [‘A’, ‘B’]

select

models.Outcomes.select(interventions)

Subset to the given interventions, preserving the iteration index.

summary

models.Outcomes.summary()

Mean of every outcome column per intervention.

Example

import pandas as pd from heormodel.models import Outcomes c = pd.DataFrame({“A”: [1.0, 3.0]}) e = pd.DataFrame({“A”: [0.1, 0.3]}) float(Outcomes.from_wide(c, e).summary()[“cost”][“A”]) 2.0