Skip to content

Data Import

The neopkpd.data module provides utilities for importing and preparing PK/PD data from various formats.

Supported Formats

  • CDISC Formats

    SDTM PC, EX, DM domains

    CDISC Import

  • NONMEM Import

    Parse NONMEM control files

    NONMEM

  • Monolix Import

    Parse Monolix project files

    Monolix

  • CSV Files

    Standard CSV data import

    CSV Import

Quick Start

Import CDISC Data

from neopkpd.data import import_cdisc

# Import from CSV files
data = import_cdisc(
    pc_path="pc.csv",      # Pharmacokinetic Concentrations
    ex_path="ex.csv",      # Exposure (Dosing)
    dm_path="dm.csv",      # Demographics
    format="csv"
)

# Access standardized data
print(f"Subjects: {len(set(data.usubjid))}")
print(f"Observations: {len(data.dv)}")

Import from XPT

from neopkpd.data import import_cdisc

data = import_cdisc(
    pc_path="pc.xpt",
    ex_path="ex.xpt",
    dm_path="dm.xpt",
    format="xpt"
)

CDISC Domains

PC (Pharmacokinetic Concentrations)

Required variables:

Variable Description Required
USUBJID Unique subject identifier Yes
PCDTC Collection datetime Yes
PCSTRESN Numeric result Yes
PCTEST Analyte name Yes
PCSPEC Specimen type No
PCBLFL Baseline flag No

EX (Exposure)

Required variables:

Variable Description Required
USUBJID Unique subject identifier Yes
EXSTDTC Start datetime Yes
EXDOSE Dose amount Yes
EXDOSU Dose units Yes
EXROUTE Route of administration No
EXDUR Duration (for infusions) No

DM (Demographics)

Common variables:

Variable Description Required
USUBJID Unique subject identifier Yes
AGE Age No
SEX Sex No
RACE Race No
WEIGHT Weight (from VS or custom) No

CSV Import

Standard Format

from neopkpd.data import read_pk_data

# Expected columns: ID, TIME, DV, AMT, EVID, CMT
data = read_pk_data("pk_data.csv")

# Custom column mapping
data = read_pk_data(
    "pk_data.csv",
    column_map={
        "subject_id": "ID",
        "time_hr": "TIME",
        "concentration": "DV",
        "dose_mg": "AMT"
    }
)

NONMEM Format

from neopkpd.data import read_nonmem_data

# Standard NONMEM dataset
data = read_nonmem_data("data.csv")

# Access as DataFrame or structured data
print(data.to_dataframe())

XPT Import

from neopkpd.data import read_xpt

# Single file
pc_data = read_xpt("pc.xpt")

# Multiple files
data = {
    "pc": read_xpt("pc.xpt"),
    "ex": read_xpt("ex.xpt"),
    "dm": read_xpt("dm.xpt")
}

Data Preparation

Time Calculation

from neopkpd.data import calculate_time_after_dose

data = calculate_time_after_dose(
    observations=obs_df,
    dosing=dose_df,
    time_column="PCDTC",
    dose_time_column="EXSTDTC"
)

BLQ Handling

from neopkpd.data import handle_blq

# Replace BLQ with LLOQ/2
data = handle_blq(
    data,
    lloq=0.1,
    method="lloq_half"
)

# Or set to zero
data = handle_blq(data, lloq=0.1, method="zero")

# Or drop
data = handle_blq(data, lloq=0.1, method="drop")

Merge Demographics

from neopkpd.data import merge_demographics

merged = merge_demographics(
    pk_data=pc_data,
    demographics=dm_data,
    on="USUBJID"
)

Data Classes

PKData

class PKData:
    ids: list[str]           # Subject IDs
    times: list[float]       # Time points
    dv: list[float]          # Observations
    amt: list[float]         # Doses
    evid: list[int]          # Event IDs
    cmt: list[int]           # Compartments
    covariates: dict         # Covariate data

    def to_dataframe(self) -> pd.DataFrame: ...
    def to_dict(self) -> dict: ...
    def subset(self, ids: list) -> PKData: ...

CDISCData

class CDISCData:
    usubjid: list[str]
    subjid: list[int]
    times: list[float]
    dv: list[float]
    doses: list[DoseEvent]
    demographics: dict

    def to_estimation_data(self) -> EstimationData: ...

Examples

Complete Workflow

from neopkpd.data import import_cdisc, handle_blq, calculate_time_after_dose
import neopkpd

# Import CDISC data
data = import_cdisc(
    pc_path="pc.csv",
    ex_path="ex.csv",
    dm_path="dm.csv"
)

# Handle BLQ
data = handle_blq(data, lloq=0.1, method="lloq_half")

# Calculate time after first dose
data = calculate_time_after_dose(data)

# Convert to estimation format
est_data = data.to_estimation_data()

# Run NCA
from neopkpd.nca import run_population_nca
nca_results = run_population_nca(est_data)

Next Steps