HTML Report Generation

Every CausalPy experiment can produce a self-contained HTML report that summarises the analysis: effect estimates, diagnostic plots, and (when used with the pipeline) sensitivity-check results. The report is ready to share with stakeholders or embed in a notebook.

import pandas as pd
import causalpy as cp

Fit an experiment

First we fit an Interrupted Time Series model — the same workflow you would use without any pipeline.

df = (
    cp.load_data("its")
    .assign(date=lambda x: pd.to_datetime(x["date"]))
    .set_index("date")
)
treatment_time = pd.to_datetime("2017-01-01")

seed = 42
model = cp.pymc_models.LinearRegression(sample_kwargs={"random_seed": seed})

result = cp.InterruptedTimeSeries(
    df,
    treatment_time,
    formula="y ~ 1 + t",
    model=model,
)

Initializing NUTS using jitter+adapt_diag...

Multiprocess sampling (4 chains in 4 jobs)

NUTS: [beta, y_hat_sigma]

Sampling 4 chains for 1_000 tune and 1_000 draw iterations (4_000 + 4_000 draws total) took 1 seconds.

Sampling: [beta, y_hat, y_hat_sigma]

Sampling: [y_hat]

Sampling: [y_hat]

Sampling: [y_hat]

Sampling: [y_hat]

Generate a report

Call .generate_report() on the fitted experiment. The method returns the HTML as a string and, optionally, writes it to a file.

html = result.generate_report()

Which we can preview below:

Show code cell source Hide code cell source

import html as html_module
import warnings

from IPython.display import HTML

with warnings.catch_warnings():
    warnings.filterwarnings(
        "ignore", "Consider using IPython.display.IFrame", UserWarning
    )
    report_widget = HTML(
        '<iframe srcdoc="' + html_module.escape(html) + '" width="100%" height="1000"'
        ' style="border: 1px solid #ddd; border-radius: 6px;"></iframe>'
    )
report_widget

Report with sensitivity checks

The standalone generate_report() covers effect estimates and plots. To include sensitivity-check results, run the analysis through a Pipeline with a SensitivityAnalysis step. The GenerateReport step then bundles everything — effect summary, diagnostic plot, and pass/fail check results — into a single report.

pipeline_result = cp.Pipeline(
    data=df,
    steps=[
        cp.EstimateEffect(
            method=cp.InterruptedTimeSeries,
            treatment_time=treatment_time,
            formula="y ~ 1 + t",
            model=cp.pymc_models.LinearRegression(sample_kwargs={"random_seed": 42}),
        ),
        cp.SensitivityAnalysis(
            checks=[
                cp.checks.PlaceboInTime(n_folds=2),
            ]
        ),
        cp.GenerateReport(include_plots=True),
    ],
).run()

Initializing NUTS using jitter+adapt_diag...

Multiprocess sampling (4 chains in 4 jobs)

NUTS: [beta, y_hat_sigma]

Sampling 4 chains for 1_000 tune and 1_000 draw iterations (4_000 + 4_000 draws total) took 1 seconds.

Sampling: [beta, y_hat, y_hat_sigma]

Sampling: [y_hat]

Sampling: [y_hat]

Sampling: [y_hat]

Sampling: [y_hat]

Fold 1 failed to fit (pseudo_treatment_time=2011-01-05 00:00:00), skipping.

Fold 2 failed to fit (pseudo_treatment_time=2014-01-03 00:00:00), skipping.

with warnings.catch_warnings():
    warnings.filterwarnings(
        "ignore", "Consider using IPython.display.IFrame", UserWarning
    )
    full_report = HTML(
        '<iframe srcdoc="'
        + html_module.escape(pipeline_result.report)
        + '" width="100%" height="1000"'
        ' style="border: 1px solid #ddd; border-radius: 6px;"></iframe>'
    )
full_report

Saving to a file

Pass output_file to write the report to disk in a single call:

result.generate_report(output_file="my_report.html")

Or save the returned string manually:

from pathlib import Path

Path("my_report.html").write_text(html, encoding="utf-8")

Note

When you use the pipeline, GenerateReport can also include sensitivity-analysis results in the report. See the pipeline notebook for that workflow.