csrlite 0.1.0__py3-none-any.whl

csrlite/ae/ae_summary.py

@@ -0,0 +1,399 @@
+# pyre-strict
+"""
+Adverse Event (AE) Analysis Functions
+
+This module provides a three-step pipeline for AE summary analysis:
+- ae_summary_ard: Generate Analysis Results Data (ARD) in long format
+- ae_summary_df: Transform ARD to wide display format
+- ae_summary_rtf: Generate formatted RTF output
+- ae_summary: Complete pipeline wrapper
+- study_plan_to_ae_summary: Batch generation from StudyPlan
+
+Uses Polars native SQL capabilities for data manipulation, count.py utilities for subject counting,
+and parse.py utilities for StudyPlan parsing.
+"""
+
+from pathlib import Path
+
+import polars as pl
+from rtflite import RTFDocument
+
+from ..common.count import count_subject, count_subject_with_observation
+from ..common.parse import StudyPlanParser
+from ..common.plan import StudyPlan
+from ..common.utils import apply_common_filters
+from .ae_utils import create_ae_rtf_table
+
+
+def study_plan_to_ae_summary(
+    study_plan: StudyPlan,
+) -> list[str]:
+    """
+    Generate AE summary RTF outputs for all analyses defined in StudyPlan.
+
+    This function reads the expanded plan from StudyPlan and generates
+    an RTF table for each analysis specification automatically.
+
+    Args:
+        study_plan: StudyPlan object with loaded datasets and analysis specifications
+
+    Returns:
+        list[str]: List of paths to generated RTF files
+    """
+
+    # Meta data
+    analysis = "ae_summary"
+    analysis_label = "Analysis of Adverse Event Summary"
+    output_dir = study_plan.output_dir
+    footnote = ["Every participant is counted a single time for each applicable row and column."]
+    source = None
+
+    population_df_name = "adsl"
+    observation_df_name = "adae"
+
+    id = ("USUBJID", "Subject ID")
+    total = True
+    missing_group = "error"
+
+    # Create output directory if it doesn't exist
+    Path(output_dir).mkdir(parents=True, exist_ok=True)
+
+    # Initialize parser
+    parser = StudyPlanParser(study_plan)
+
+    # Get expanded plan DataFrame
+    plan_df = study_plan.get_plan_df()
+
+    # Filter for AE summary analyses
+    ae_plans = plan_df.filter(pl.col("analysis") == analysis)
+
+    rtf_files = []
+
+    # Generate RTF for each analysis
+    for row in ae_plans.iter_rows(named=True):
+        population = row["population"]
+        observation = row.get("observation")
+        parameter = row["parameter"]
+        group = row.get("group")
+
+        # Validate group is specified
+        if group is None:
+            raise ValueError(
+                f"Group not specified in YAML "
+                f"population={population}, observation={observation}, parameter={parameter}. "
+                "Please add group to your YAML plan."
+            )
+
+        # Get datasets using parser
+        population_df, observation_df = parser.get_datasets(population_df_name, observation_df_name)
+
+        # Get filters and configuration using parser
+        population_filter = parser.get_population_filter(population)
+        param_names, param_filters, param_labels, _ = parser.get_parameter_info(
+            parameter
+        )  # Ignore indent for AE
+        obs_filter = parser.get_observation_filter(observation)
+        group_var_name, group_labels = parser.get_group_info(group)
+
+        # Build variables as list of tuples [(filter, label)]
+        variables_list = list(zip(param_filters, param_labels))
+
+        # Build group tuple (variable_name, label)
+        group_var_label = group_labels[0] if group_labels else group_var_name
+        group_tuple = (group_var_name, group_var_label)
+
+        # Build title with population and observation context
+        title_parts = [analysis_label]
+        if observation:
+            obs_kw = study_plan.keywords.observations.get(observation)
+            if obs_kw and obs_kw.label:
+                title_parts.append(obs_kw.label)
+
+        pop_kw = study_plan.keywords.populations.get(population)
+        if pop_kw and pop_kw.label:
+            title_parts.append(pop_kw.label)
+
+        # Build output filename
+        filename = f"{analysis}_{population}"
+        if observation:
+            filename += f"_{observation}"
+        filename += f"_{parameter.replace(';', '_')}.rtf"
+        output_file = str(Path(output_dir) / filename)
+
+        # Generate RTF using the new ae_summary signature
+        rtf_path = ae_summary(
+            population=population_df,
+            observation=observation_df,
+            population_filter=population_filter,
+            observation_filter=obs_filter,
+            id=id,
+            group=group_tuple,
+            variables=variables_list,
+            title=title_parts,
+            footnote=footnote,
+            source=source,
+            output_file=output_file,
+            total=total,
+            missing_group=missing_group,
+        )
+
+        rtf_files.append(rtf_path)
+
+    return rtf_files
+
+
+def ae_summary(
+    population: pl.DataFrame,
+    observation: pl.DataFrame,
+    population_filter: str | None,
+    observation_filter: str | None,
+    id: tuple[str, str],
+    group: tuple[str, str],
+    variables: list[tuple[str, str]],
+    title: list[str],
+    footnote: list[str] | None,
+    source: list[str] | None,
+    output_file: str,
+    total: bool = True,
+    col_rel_width: list[float] | None = None,
+    missing_group: str = "error",
+) -> str:
+    """
+    Complete AE summary pipeline wrapper.
+
+    This function orchestrates the three-step pipeline:
+    1. ae_summary_ard: Generate Analysis Results Data
+    2. ae_summary_df: Transform to display format
+    3. ae_summary_rtf: Generate RTF output and write to file
+
+    Args:
+        population: Population DataFrame (subject-level data, e.g., ADSL)
+        observation: Observation DataFrame (event data, e.g., ADAE)
+        population_filter: SQL WHERE clause for population (can be None)
+        observation_filter: SQL WHERE clause for observation (can be None)
+        id: Tuple (variable_name, label) for ID column
+        group: Tuple (variable_name, label) for grouping variable
+        variables: List of tuples [(filter, label)] for analysis variables
+        title: Title for RTF output as list of strings
+        footnote: Optional footnote for RTF output as list of strings
+        source: Optional source for RTF output as list of strings
+        output_file: File path to write RTF output
+        total: Whether to include total column (default: True)
+        col_rel_width: Optional column widths for RTF output
+        missing_group: How to handle missing group values (default: "error")
+
+    Returns:
+        str: Path to the generated RTF file
+    """
+    # Step 1: Generate ARD
+    ard = ae_summary_ard(
+        population=population,
+        observation=observation,
+        population_filter=population_filter,
+        observation_filter=observation_filter,
+        id=id,
+        group=group,
+        variables=variables,
+        total=total,
+        missing_group=missing_group,
+    )
+
+    # Step 2: Transform to display format
+    df = ae_summary_df(ard)
+
+    # Step 3: Generate RTF and write to file
+    rtf_doc = ae_summary_rtf(
+        df=df,
+        title=title,
+        footnote=footnote,
+        source=source,
+        col_rel_width=col_rel_width,
+    )
+    rtf_doc.write_rtf(output_file)
+
+    return output_file
+
+
+def ae_summary_ard(
+    population: pl.DataFrame,
+    observation: pl.DataFrame,
+    population_filter: str | None,
+    observation_filter: str | None,
+    id: tuple[str, str],
+    group: tuple[str, str],
+    variables: list[tuple[str, str]],
+    total: bool,
+    missing_group: str,
+) -> pl.DataFrame:
+    """
+    Generate Analysis Results Data (ARD) for AE summary analysis.
+
+    Creates a long-format DataFrame with standardized structure (__index__, __group__, __value__)
+    containing population counts and observation statistics for each analysis variable.
+
+    Args:
+        population: Population DataFrame (subject-level data, e.g., ADSL)
+        observation: Observation DataFrame (event data, e.g., ADAE)
+        population_filter: SQL WHERE clause for population (can be None)
+        observation_filter: SQL WHERE clause for observation (can be None)
+        id: Tuple (variable_name, label) for ID column
+        group: Tuple (variable_name, label) for grouping variable
+        variables: List of tuples [(filter, label)] for analysis variables
+        total: Whether to include total column in counts
+        missing_group: How to handle missing group values: "error", "ignore", or "fill"
+
+    Returns:
+        pl.DataFrame: Long-format ARD with columns __index__, __group__, __value__
+    """
+    # Extract group variable name (label is in tuple but not needed separately)
+    pop_var_name = "Participants in population"
+    id_var_name, id_var_label = id
+    group_var_name, group_var_label = group
+
+    # Apply common filters (parameter_filter is handled inside the loop, so None here)
+    population_filtered, observation_to_filter = apply_common_filters(
+        population=population,
+        observation=observation,
+        population_filter=population_filter,
+        observation_filter=observation_filter,
+    )
+
+    # Filter observation data to include only subjects in the filtered population
+    # Process all variables in the list
+    observation_filtered_list = []
+    for variable_filter, variable_label in variables:
+        obs_filtered = (
+            observation_to_filter.filter(
+                pl.col(id_var_name).is_in(population_filtered[id_var_name].to_list())
+            )
+            .filter(pl.sql_expr(variable_filter))
+            .with_columns(pl.lit(variable_label).alias("__index__"))
+        )
+
+        observation_filtered_list.append(obs_filtered)
+
+    # Concatenate all filtered observations
+    observation_filtered = pl.concat(observation_filtered_list)
+
+    # Population
+    n_pop = count_subject(
+        population=population_filtered,
+        id=id_var_name,
+        group=group_var_name,
+        total=total,
+        missing_group=missing_group,
+    )
+
+    n_pop = n_pop.select(
+        pl.lit(pop_var_name).alias("__index__"),
+        pl.col(group_var_name).alias("__group__"),
+        pl.col("n_subj_pop").cast(pl.String).alias("__value__"),
+    )
+
+    # Empty row with same structure as n_pop but with empty strings
+    n_empty = n_pop.select(
+        pl.lit("").alias("__index__"), pl.col("__group__"), pl.lit("").alias("__value__")
+    )
+
+    # Observation
+    n_obs = count_subject_with_observation(
+        population=population_filtered,
+        observation=observation_filtered,
+        id=id_var_name,
+        group=group_var_name,
+        total=total,
+        variable="__index__",
+        missing_group=missing_group,
+    )
+
+    n_obs = n_obs.select(
+        pl.col("__index__"),
+        pl.col(group_var_name).alias("__group__"),
+        pl.col("n_pct_subj_fmt").alias("__value__"),
+    )
+
+    res = pl.concat([n_pop, n_empty, n_obs])
+
+    # Convert __index__ to ordered Enum based on appearance
+    # Build the ordered categories list: population name, empty string, then variable labels
+    variable_labels = [label for _, label in variables]
+    ordered_categories = [pop_var_name, ""] + variable_labels
+
+    res = res.with_columns(pl.col("__index__").cast(pl.Enum(ordered_categories))).sort(
+        "__index__", "__group__"
+    )
+
+    return res
+
+
+def ae_summary_df(ard: pl.DataFrame) -> pl.DataFrame:
+    """
+    Transform AE summary ARD (Analysis Results Data) into display-ready DataFrame.
+
+    Converts the long-format ARD with __index__, __group__, and __value__ columns
+    into a wide-format display table where groups become columns.
+
+    Args:
+        ard: Analysis Results Data DataFrame with __index__, __group__, __value__ columns
+
+    Returns:
+        pl.DataFrame: Wide-format display table with groups as columns
+    """
+    # Pivot from long to wide format: __group__ values become columns
+    df_wide = ard.pivot(index="__index__", on="__group__", values="__value__")
+
+    return df_wide
+
+
+def ae_summary_rtf(
+    df: pl.DataFrame,
+    title: list[str],
+    footnote: list[str] | None,
+    source: list[str] | None,
+    col_rel_width: list[float] | None = None,
+) -> RTFDocument:
+    """
+    Generate RTF table from AE summary display DataFrame.
+
+    Creates a formatted RTF table with two-level column headers showing
+    treatment groups with "n (%)" values.
+
+    Args:
+        df: Display DataFrame from ae_summary_df (wide format with __index__ column)
+        title: Title(s) for the table as list of strings
+        footnote: Optional footnote(s) as list of strings
+        source: Optional source note(s) as list of strings
+        col_rel_width: Optional list of relative column widths. If None, auto-calculated
+                       as [n_cols-1, 1, 1, 1, ...] where n_cols is total column count
+
+    Returns:
+        RTFDocument: RTF document object that can be written to file
+    """
+
+    # Rename __index__ to empty string for display
+    df_rtf = df.rename({"__index__": ""})
+
+    # Calculate number of columns
+    n_cols = len(df_rtf.columns)
+
+    # Build first-level column headers (use actual column names)
+    col_header_1 = list(df_rtf.columns)
+
+    # Build second-level column headers (empty for first, "n (%)" for groups)
+    col_header_2 = [""] + ["n (%)"] * (n_cols - 1)
+
+    # Calculate column widths - auto-calculate if not provided
+    if col_rel_width is None:
+        col_widths = [float(n_cols - 1)] + [1.0] * (n_cols - 1)
+    else:
+        col_widths = col_rel_width
+
+    return create_ae_rtf_table(
+        df=df_rtf,
+        col_header_1=col_header_1,
+        col_header_2=col_header_2,
+        col_widths=col_widths,
+        title=title,
+        footnote=footnote,
+        source=source,
+    )

csrlite/ae/ae_utils.py

@@ -0,0 +1,132 @@
+# pyre-strict
+from typing import Any
+
+import polars as pl
+from rtflite import RTFBody, RTFColumnHeader, RTFDocument, RTFFootnote, RTFPage, RTFSource, RTFTitle
+
+
+def get_ae_parameter_title(param: Any, prefix: str = "Participants With") -> str:
+    """
+    Extract title from parameter for ae_* title generation.
+
+    Args:
+        param: Parameter object with terms attribute
+        prefix: Prefix for the title (e.g. "Participants With", "Listing of Participants With")
+
+    Returns:
+        Title string for the analysis
+    """
+    default_suffix = "Adverse Events"
+
+    if not param:
+        return f"{prefix} {default_suffix}"
+
+    # Check for terms attribute
+    if hasattr(param, "terms") and param.terms and isinstance(param.terms, dict):
+        terms = param.terms
+
+        # Preprocess to empty strings (avoiding None)
+        before = terms.get("before", "").title()
+        after = terms.get("after", "").title()
+
+        # Build title and clean up extra spaces
+        title = f"{prefix} {before} {default_suffix} {after}"
+        return " ".join(title.split())  # Remove extra spaces
+
+    # Fallback to default
+    return f"{prefix} {default_suffix}"
+
+
+def get_ae_parameter_row_labels(param: Any) -> tuple[str, str]:
+    """
+    Generate n_with and n_without row labels based on parameter terms.
+
+    Returns:
+        Tuple of (n_with_label, n_without_label)
+    """
+    # Default labels
+    default_with = "    with one or more adverse events"
+    default_without = "    with no adverse events"
+
+    if not param or not hasattr(param, "terms") or not param.terms:
+        return (default_with, default_without)
+
+    terms = param.terms
+    before = terms.get("before", "").lower()
+    after = terms.get("after", "").lower()
+
+    # Build dynamic labels with leading indentation
+    with_label = f"with one or more {before} adverse events {after}"
+    without_label = f"with no {before} adverse events {after}"
+
+    # Clean up extra spaces and add back the 4-space indentation
+    with_label = "    " + " ".join(with_label.split())
+    without_label = "    " + " ".join(without_label.split())
+
+    return (with_label, without_label)
+
+
+def create_ae_rtf_table(
+    df: pl.DataFrame,
+    col_header_1: list[str],
+    col_header_2: list[str] | None,
+    col_widths: list[float] | None,
+    title: list[str] | str,
+    footnote: list[str] | str | None,
+    source: list[str] | str | None,
+    borders_2: bool = True,
+    orientation: str = "landscape",
+) -> RTFDocument:
+    """
+    Create a standardized RTF table document with 1 or 2 header rows.
+    """
+    n_cols = len(df.columns)
+
+    # Calculate column widths if None - simple default
+    if col_widths is None:
+        col_widths = [1] * n_cols
+
+    # Normalize metadata
+    title_list = [title] if isinstance(title, str) else title
+    footnote_list = [footnote] if isinstance(footnote, str) else (footnote or [])
+    source_list = [source] if isinstance(source, str) else (source or [])
+
+    headers = [
+        RTFColumnHeader(
+            text=col_header_1,
+            col_rel_width=col_widths,
+            text_justification=["l"] + ["c"] * (n_cols - 1),
+        )
+    ]
+
+    if col_header_2:
+        h2_kwargs = {
+            "text": col_header_2,
+            "col_rel_width": col_widths,
+            "text_justification": ["l"] + ["c"] * (n_cols - 1),
+        }
+        if borders_2:
+            h2_kwargs["border_left"] = ["single"]
+            h2_kwargs["border_top"] = [""]
+
+        headers.append(RTFColumnHeader(**h2_kwargs))
+
+    rtf_components: dict[str, Any] = {
+        "df": df,
+        "rtf_page": RTFPage(orientation=orientation),
+        "rtf_title": RTFTitle(text=title_list),
+        "rtf_column_header": headers,
+        "rtf_body": RTFBody(
+            col_rel_width=col_widths,
+            text_justification=["l"] + ["c"] * (n_cols - 1),
+            border_left=["single"] * n_cols,
+        ),
+    }
+
+    if footnote_list:
+        rtf_components["rtf_footnote"] = RTFFootnote(text=footnote_list)
+
+    if source_list:
+        rtf_components["rtf_source"] = RTFSource(text=source_list)
+
+    return RTFDocument(**rtf_components)