Demo for the DoWhy causal API

We show a simple example of adding a causal extension to any dataframe.

import dowhy
dowhy.enable_notebook_rendering()

import dowhy.datasets
import dowhy.api
from dowhy.graph import build_graph_from_str

import numpy as np
import pandas as pd

from statsmodels.api import OLS

data = dowhy.datasets.linear_dataset(beta=5,
        num_common_causes=1,
        num_instruments = 0,
        num_samples=1000,
        treatment_is_binary=True)
df = data['df']
df['y'] = df['y'] + np.random.normal(size=len(df)) # Adding noise to data. Without noise, the variance in Y|X, Z is zero, and mcmc fails.
nx_graph = build_graph_from_str(data["dot_graph"])

treatment= data["treatment_name"][0]
outcome = data["outcome_name"][0]
common_cause = data["common_causes_names"][0]
df

	W0	v0	y
0	-0.100021	False	0.132437
1	0.534403	True	4.776436
2	0.111168	False	-0.824853
3	1.370331	True	8.595711
4	-0.618126	False	-0.541416
...	...	...	...
995	-0.379531	True	2.793668
996	-1.785375	True	2.363306
997	-2.354834	True	1.268439
998	-0.957736	True	3.688295
999	-1.240677	False	-0.713537

1000 rows × 3 columns

# data['df'] is just a regular pandas.DataFrame
df.causal.do(x=treatment,
             variable_types={treatment: 'b', outcome: 'c', common_cause: 'c'},
             outcome=outcome,
             common_causes=[common_cause],
            ).groupby(treatment).mean().plot(y=outcome, kind='bar')

<Axes: xlabel='v0'>

df.causal.do(x={treatment: 1},
              variable_types={treatment:'b', outcome: 'c', common_cause: 'c'},
              outcome=outcome,
              method='weighting',
              common_causes=[common_cause]
              ).groupby(treatment).mean().plot(y=outcome, kind='bar')

<Axes: xlabel='v0'>

cdf_1 = df.causal.do(x={treatment: 1},
              variable_types={treatment: 'b', outcome: 'c', common_cause: 'c'},
              outcome=outcome,
              graph=nx_graph
              )

cdf_0 = df.causal.do(x={treatment: 0},
              variable_types={treatment: 'b', outcome: 'c', common_cause: 'c'},
              outcome=outcome,
              graph=nx_graph
              )

cdf_0

	W0	v0	y	propensity_score	weight
0	-0.279066	False	0.311358	0.504760	1.981141
1	-0.355988	False	-0.875442	0.507923	1.968801
2	1.127934	False	2.253372	0.447079	2.236739
3	0.929628	False	1.149448	0.455159	2.197035
4	1.048707	False	2.814462	0.450304	2.220721
...	...	...	...	...	...
995	1.967333	False	3.934154	0.413237	2.419917
996	-0.929117	False	-2.203187	0.531459	1.881612
997	-1.862456	False	-2.065187	0.569443	1.756101
998	-0.014084	False	0.942722	0.493860	2.024867
999	0.373045	False	0.639678	0.477948	2.092276

1000 rows × 5 columns

cdf_1

	W0	v0	y	propensity_score	weight
0	-1.016380	True	2.974043	0.464967	2.150690
1	-1.750903	True	3.555488	0.435063	2.298520
2	-0.605960	True	2.955557	0.481801	2.075546
3	-0.125322	True	4.851993	0.501565	1.993761
4	-1.366810	True	2.508822	0.450655	2.218991
...	...	...	...	...	...
995	0.318000	True	8.975365	0.519791	1.923849
996	1.241249	True	5.593692	0.557525	1.793641
997	-1.381995	True	3.793374	0.450037	2.222041
998	0.801727	True	7.344791	0.539617	1.853165
999	-1.030269	True	3.333833	0.464399	2.153323

1000 rows × 5 columns

Comparing the estimate to Linear Regression

First, estimating the effect using the causal data frame, and the 95% confidence interval.

(cdf_1['y'] - cdf_0['y']).mean()

$\displaystyle 5.06699896618239$

1.96*(cdf_1['y'] - cdf_0['y']).std() / np.sqrt(len(df))

$\displaystyle 0.163468403376322$

Comparing to the estimate from OLS.

model = OLS(np.asarray(df[outcome]), np.asarray(df[[common_cause, treatment]], dtype=np.float64))
result = model.fit()
result.summary()

OLS Regression Results

Dep. Variable:	y	R-squared (uncentered):	0.935
Model:	OLS	Adj. R-squared (uncentered):	0.935
Method:	Least Squares	F-statistic:	7149.
Date:	Wed, 22 Apr 2026	Prob (F-statistic):	0.00
Time:	08:22:17	Log-Likelihood:	-1405.6
No. Observations:	1000	AIC:	2815.
Df Residuals:	998	BIC:	2825.
Df Model:	2
Covariance Type:	nonrobust

	coef	std err	t	P>|t|	[0.025	0.975]
x1	1.6382	0.030	54.469	0.000	1.579	1.697
x2	5.0559	0.045	112.840	0.000	4.968	5.144

Omnibus:	0.105	Durbin-Watson:	1.958
Prob(Omnibus):	0.949	Jarque-Bera (JB):	0.176
Skew:	-0.005	Prob(JB):	0.916
Kurtosis:	2.936	Cond. No.	1.53

Notes:
[1] R² is computed without centering (uncentered) since the model does not contain a constant.
[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.