admiralpy

admiralpy is a Python implementation of the admiral R package — an open-source toolbox for creating ADaM (Analysis Data Model) datasets for clinical trials following CDISC standards.

New to admiralpy?

Start with the Get Started guide to see working examples of the most common derivation workflows.

Installation

pip install admiralpy

Or install directly from the GitHub repository:

pip install git+https://github.com/bms63/admiralpy.git

Available Functions

Category	Function	Description
Date/Time	`derive_vars_dt()`	Derive a date variable (`*DT`) from a DTC character variable
	`derive_vars_dtm()`	Derive a datetime variable (`*DTM`) from a DTC character variable
Baseline	`derive_var_base()`	Derive a baseline variable (e.g. `BASE`)
	`derive_var_chg()`	Derive change from baseline (`CHG`)
	`derive_var_pchg()`	Derive percent change from baseline (`PCHG`)
Merge	`derive_vars_merged()`	Add variables from an additional dataset
Parameters	`derive_param_computed()`	Add a derived parameter record (e.g. MAP)
Filter	`filter_extreme()`	Filter first/last record per by-group
Convert	`convert_dtc_to_dt()`	Convert DTC string to a date
	`convert_dtc_to_dtm()`	Convert DTC string to a datetime
Compute	`compute_age_years()`	Convert age to years
	`compute_bmi()`	Body Mass Index
	`compute_bsa()`	Body Surface Area (4 formulae)

Design Principles

admiralpy follows the same design philosophy as the R admiral package:

Usability — thoroughly documented functions with clear examples.
Simplicity — each function has a single, well-defined purpose.
Findability — consistent naming: derive_vars_*, derive_var_*, derive_param_*, compute_*, convert_*, filter_*.
Readability — chain derivations together with standard Python pandas method chaining.

Quick Example

import pandas as pd
from admiralpy import (
    derive_vars_dt,
    derive_var_base,
    derive_var_chg,
    derive_var_pchg,
    compute_bmi,
)

# Convert DTC columns to date variables
mhdt = pd.DataFrame({
    "MHSTDTC": ["2019-07-18T15:25:40", "2019-07-18", "2019-02", "2019", ""]
})
mhdt = derive_vars_dt(
    mhdt,
    new_vars_prefix="AST",
    dtc="MHSTDTC",
    highest_imputation="M",
    date_imputation="first",
)

# Derive baseline, change and percent change
advs = pd.DataFrame({
    "USUBJID": ["P01", "P01", "P01"],
    "PARAMCD": ["WEIGHT"] * 3,
    "AVAL":    [80.0, 80.8, 81.4],
    "ABLFL":   ["Y", None, None],
})
advs = (advs
    .pipe(derive_var_base, by_vars=["USUBJID", "PARAMCD"])
    .pipe(derive_var_chg)
    .pipe(derive_var_pchg)
)