TimeRangePhenotype

Bases: Phenotype

As the name implies, TimeRangePhenotype is designed for working with time ranges. If the input data has a start and an end date, use TimeRangePhenotype to identify other events (or patients) that occur within this time range. The most common use case of this is working with 'health insurance coverage' data i.e. on 'OBSERVATION_PERIOD' table. These tables have one or many rows per patient with the start of coverage and end of coverage i.e. domains compatible with TimeRangePhenotype require a START_DATE and an END_DATE column. At it's simplest, TimeRangePhenotype identifies patients who have their INDEX_DATE (or other anchor date of interest) within this time range. Additionally, a minimum or maximum number of days from the anchor date to the beginning/end of the time range can be defined. The returned Phenotype has the following interpretation:

DATE: If relative_time_range.when='before', then DATE is the beginning of the coverage period containing the anchor phenotype. If relative_time_range.when='after', then DATE is the end of the coverage period containing the anchor date. VALUE: Coverage (in days) relative to the anchor date. By convention, always non-negative.

There are two primary use cases for TimeRangePhenotype

1. Identify patients with some minimum duration of coverage prior to anchor_phenotype date e.g. "identify patients with 1 year of continuous coverage prior to index date"
2. Determine the date of loss to followup (right censoring) i.e. the duration of coverage after the anchor_phenotype event

Data for TimeRangePhenotype

This phenotype requires a table with PersonID and a coverage start date and end date. Depending on the datasource used, this information is a separate ObservationPeriod table or found in the PersonTable. Use an PhenexObservationPeriodTable to map required coverage start and end date columns. For tables with overlapping time ranges, use the CombineOverlappingPeriods derived table to combine time ranges into a single time range.

PersonID	startDate	endDate
1	2009-01-01	2010-01-01
2	2008-01-01	2010-01-02

One assumption that is made by TimeRangePhenotype is that there are NO overlapping coverage periods.

Parameters:

Name	Type	Description	Default
name	Optional[str]	The name of the phenotype.	'TIME_RANGE'
domain	Optional[str]	The domain of the phenotype. Default is 'observation_period'.	'OBSERVATION_PERIOD'
relative_time_range	Optional[RelativeTimeRangeFilter]	Filter returned persons based on the duration of coverage in days. The relative_time_range.anchor_phenotype defines the reference date with respect to calculate coverage. In typical applications, the anchor phenotype will be the entry criterion. The relative_time_range.when 'before', 'after'. If before, the return date is the start of the coverage period containing the anchor_phenotype. If after, the return date is the end of the coverage period containing the anchor_phenotype.	None
allow_null_end_date	bool	TimeRangePhenotype checks that anchor date is within the time range of interest. This requires that the start date is not null, and the end date is either null or after the anchor date. If you want to require that the end date is not null, set allow_null_end_date to False.	True

Example:

# make sure to create an entry phenotype, for example 'atrial fibrillation diagnosis'
entry_phenotype = CodelistPhenotype(...)
# one year continuous coverage prior to index
one_year_coverage = TimeRangePhenotype(
    relative_time_range = RelativeTimeRangeFilter(
        min_days=GreaterThanOrEqualTo(365),
        anchor_phenotype = entry_phenotype,
        when = 'before',
    ),
)
# determine the date of loss to followup
loss_to_followup = TimeRangePhenotype(
    relative_time_range = RelativeTimeRangeFilter(
        anchor_phenotype = entry_phenotype
        when = 'after',
    )
)

# determine the date when a drug was discontinued
drug_discontinuation = TimeRangePhenotype(
    relative_time_range = RelativeTimeRangeFilter(
        anchor_phenotype = entry_phenotype
        when = 'after',
    )
)

Source code in phenex/phenotypes/time_range_phenotype.py

class TimeRangePhenotype(Phenotype):
    """
    As the name implies, TimeRangePhenotype is designed for working with time ranges. If the input data has a start and an end date, use TimeRangePhenotype to identify other events (or patients) that occur within this time range. The most common use case of this is working with 'health insurance coverage' data i.e. on 'OBSERVATION_PERIOD' table. These tables have one or many rows per patient with the start of coverage and end of coverage i.e. domains compatible with TimeRangePhenotype require a START_DATE and an END_DATE column. At it's simplest, TimeRangePhenotype identifies patients who have their INDEX_DATE (or other anchor date of interest) within this time range. Additionally, a minimum or maximum number of days from the anchor date to the beginning/end of the time range can be defined. The returned Phenotype has the following interpretation:

    DATE: If relative_time_range.when='before', then DATE is the beginning of the coverage period containing the anchor phenotype. If relative_time_range.when='after', then DATE is the end of the coverage period containing the anchor date.
    VALUE: Coverage (in days) relative to the anchor date. By convention, always non-negative.

    There are two primary use cases for TimeRangePhenotype:
        1. Identify patients with some minimum duration of coverage prior to anchor_phenotype date e.g. "identify patients with 1 year of continuous coverage prior to index date"
        2. Determine the date of loss to followup (right censoring) i.e. the duration of coverage after the anchor_phenotype event

    ## Data for TimeRangePhenotype
    This phenotype requires a table with PersonID and a coverage start date and end date. Depending on the datasource used, this information is a separate ObservationPeriod table or found in the PersonTable. Use an PhenexObservationPeriodTable to map required coverage start and end date columns. For tables with overlapping time ranges, use the CombineOverlappingPeriods derived table to combine time ranges into a single time range.

    | PersonID    |   startDate          |   endDate          |
    |-------------|----------------------|--------------------|
    | 1           |   2009-01-01         |   2010-01-01       |
    | 2           |   2008-01-01         |   2010-01-02       |

    One assumption that is made by TimeRangePhenotype is that there are **NO overlapping coverage periods**.

    Parameters:
        name: The name of the phenotype.
        domain: The domain of the phenotype. Default is 'observation_period'.
        relative_time_range: Filter returned persons based on the duration of coverage in days. The relative_time_range.anchor_phenotype defines the reference date with respect to calculate coverage. In typical applications, the anchor phenotype will be the entry criterion. The relative_time_range.when 'before', 'after'. If before, the return date is the start of the coverage period containing the anchor_phenotype. If after, the return date is the end of the coverage period containing the anchor_phenotype.
        allow_null_end_date: TimeRangePhenotype checks that anchor date is within the time range of interest. This requires that the start date is not null, and the end date is either null or after the anchor date. If you want to require that the end date is not null, set allow_null_end_date to False.

    Example:
    ```python
    # make sure to create an entry phenotype, for example 'atrial fibrillation diagnosis'
    entry_phenotype = CodelistPhenotype(...)
    # one year continuous coverage prior to index
    one_year_coverage = TimeRangePhenotype(
        relative_time_range = RelativeTimeRangeFilter(
            min_days=GreaterThanOrEqualTo(365),
            anchor_phenotype = entry_phenotype,
            when = 'before',
        ),
    )
    # determine the date of loss to followup
    loss_to_followup = TimeRangePhenotype(
        relative_time_range = RelativeTimeRangeFilter(
            anchor_phenotype = entry_phenotype
            when = 'after',
        )
    )

    # determine the date when a drug was discontinued
    drug_discontinuation = TimeRangePhenotype(
        relative_time_range = RelativeTimeRangeFilter(
            anchor_phenotype = entry_phenotype
            when = 'after',
        )
    )
    ```
    """

    def __init__(
        self,
        name: Optional[str] = "TIME_RANGE",
        domain: Optional[str] = "OBSERVATION_PERIOD",
        relative_time_range: Optional["RelativeTimeRangeFilter"] = None,
        allow_null_end_date: bool = True,
        **kwargs
    ):
        super(TimeRangePhenotype, self).__init__(name=name, **kwargs)
        self.domain = domain
        self.relative_time_range = relative_time_range
        self.allow_null_end_date = allow_null_end_date
        if self.relative_time_range is not None:
            if self.relative_time_range.anchor_phenotype is not None:
                self.add_children(self.relative_time_range.anchor_phenotype)

    def _execute(self, tables: Dict[str, Table]) -> PhenotypeTable:
        table = tables[self.domain]
        table, reference_column = attach_anchor_and_get_reference_date(
            table, self.relative_time_range.anchor_phenotype
        )

        # Ensure that the observation period includes anchor date
        # Allow END_DATE to be null (ongoing periods) if allow_null_end_date is True
        if self.allow_null_end_date:
            table = table.filter(
                (table.START_DATE <= reference_column)
                & ((reference_column <= table.END_DATE) | (table.END_DATE.isnull()))
            )
        else:
            table = table.filter(
                (table.START_DATE <= reference_column)
                & (reference_column <= table.END_DATE)
            )

        if (
            self.relative_time_range is None
            or self.relative_time_range.when == "before"
        ):
            VALUE = reference_column.delta(table.START_DATE, "day")
            EVENT_DATE = table.START_DATE
        else:
            VALUE = table.END_DATE.delta(reference_column, "day")
            EVENT_DATE = table.END_DATE

        table = table.mutate(VALUE=VALUE, EVENT_DATE=EVENT_DATE)

        if self.relative_time_range is not None:
            value_filter = ValueFilter(
                min_value=self.relative_time_range.min_days,
                max_value=self.relative_time_range.max_days,
                column_name="VALUE",
            )
            ibis.options.interactive = True
            table = value_filter.filter(table)

        return self._perform_final_processing(table)

dependencies property

Recursively collect all dependencies of a node (including dependencies of dependencies).

Returns:

Type	Description
Set[Node]	List[Node]: A list of Node objects on which this Node depends.

dependency_graph property

Build a dependency graph where each node maps to its direct dependencies (children).

Returns:

Type	Description
Dict[Node, Set[Node]]	Dict[Node, Set[Node]: A mapping of Node's to their children Node's.

execution_metadata property

Retrieve the full execution metadata row for this node from the local DuckDB database.

Returns:

Type	Description
	pandas.Series: A series containing NODE_NAME, LAST_HASH, NODE_PARAMS, and LAST_EXECUTED for this node, or None if the node has never been executed.

namespaced_table property

A PhenotypeTable has generic column names 'person_id', 'boolean', 'event_date', and 'value'. The namespaced_table prepends the phenotype name to all of these columns. This is useful when joining multiple phenotype tables together.

Returns:

Name	Type	Description
table	Table	The namespaced table for the current phenotype.

reverse_dependency_graph property

Build a reverse dependency graph where each node maps to nodes that depend on it (parents).

Returns:

Type	Description
Dict[Node, Set[Node]]	Dict[Node, List[Node]: A mapping of Node's to their parent Node's.

clear_cache(con=None, recursive=False)

Clear the cached state for this node, forcing re-execution on the next call to execute().

This method removes the node's hash from the node states table and optionally drops the materialized table from the database. After calling this method, the node will be treated as if it has never been executed before.

Parameters:

Name	Type	Description	Default
con	Optional[object]	Database connector. If provided, will also drop the materialized table from the database.	None
recursive	bool	If True, also clear the cache for all child nodes recursively. Defaults to False.	False

Example

# Clear cache for a single node
my_node.clear_cache()

# Clear cache and drop materialized table
my_node.clear_cache(con=my_connector)

# Clear cache for node and all its dependencies
my_node.clear_cache(recursive=True)

Source code in phenex/node.py

def clear_cache(self, con: Optional[object] = None, recursive: bool = False):
    """
    Clear the cached state for this node, forcing re-execution on the next call to execute().

    This method removes the node's hash from the node states table and optionally drops the materialized table from the database. After calling this method, the node will be treated as if it has never been executed before.

    Parameters:
        con: Database connector. If provided, will also drop the materialized table from the database.
        recursive: If True, also clear the cache for all child nodes recursively. Defaults to False.

    Example:
        ```python
        # Clear cache for a single node
        my_node.clear_cache()

        # Clear cache and drop materialized table
        my_node.clear_cache(con=my_connector)

        # Clear cache for node and all its dependencies
        my_node.clear_cache(recursive=True)
        ```
    """
    logger.info(f"Node '{self.name}': clearing cached state...")

    # Clear the hash from the node states table
    with Node._hash_update_lock:
        duckdb_con = DuckDBConnector(DUCKDB_DEST_DATABASE=NODE_STATES_DB_NAME)
        if NODE_STATES_TABLE_NAME in duckdb_con.dest_connection.list_tables():
            table = duckdb_con.get_dest_table(NODE_STATES_TABLE_NAME).to_pandas()
            # Remove this node's entry
            table = table[table.NODE_NAME != self.name]

            # Update the table
            if len(table) > 0:
                updated_table = ibis.memtable(table)
                duckdb_con.create_table(
                    updated_table, name_table=NODE_STATES_TABLE_NAME, overwrite=True
                )
            else:
                # Drop the table if it's empty
                duckdb_con.dest_connection.drop_table(NODE_STATES_TABLE_NAME)

    # Drop materialized table if connector is provided
    if con is not None:
        try:
            if self.name in con.dest_connection.list_tables():
                logger.info(f"Node '{self.name}': dropping materialized table...")
                con.dest_connection.drop_table(self.name)
        except Exception as e:
            logger.warning(
                f"Node '{self.name}': failed to drop materialized table: {e}"
            )

    # Reset the table attribute
    self.table = None

    # Recursively clear children if requested
    if recursive:
        for child in self.children:
            child.clear_cache(con=con, recursive=recursive)

    logger.info(f"Node '{self.name}': cache cleared successfully.")

execute(tables=None, con=None, overwrite=False, lazy_execution=False, n_threads=1)

Executes the Node computation for the current node and its dependencies. Supports lazy execution using hash-based change detection to avoid recomputing Node's that have already executed.

Parameters:

Name	Type	Description	Default
tables	Dict[str, Table]	A dictionary mapping domains to Table objects.	None
con	Optional[object]	Connection to database for materializing outputs. If provided, outputs from the node and all children nodes will be materialized (written) to the database using the connector.	None
overwrite	bool	If True, will overwrite any existing tables found in the database while writing. If False, will throw an error when an existing table is found. Has no effect if con is not passed.	False
lazy_execution	bool	If True, only re-executes if the node's definition has changed. Defaults to False. You should pass overwrite=True with lazy_execution as lazy_execution is intended precisely for iterative updates to a node definition. You must pass a connector (to cache results) for lazy_execution to work.	False
n_threads	int	Max number of Node's to execute simultaneously when this node has multiple children.	1

Returns:

Name	Type	Description
Table	Table	The resulting table for this node. Also accessible through self.table after calling self.execute().

Source code in phenex/node.py

def execute(
    self,
    tables: Dict[str, Table] = None,
    con: Optional[object] = None,
    overwrite: bool = False,
    lazy_execution: bool = False,
    n_threads: int = 1,
) -> Table:
    """
    Executes the Node computation for the current node and its dependencies. Supports lazy execution using hash-based change detection to avoid recomputing Node's that have already executed.

    Parameters:
        tables: A dictionary mapping domains to Table objects.
        con: Connection to database for materializing outputs. If provided, outputs from the node and all children nodes will be materialized (written) to the database using the connector.
        overwrite: If True, will overwrite any existing tables found in the database while writing. If False, will throw an error when an existing table is found. Has no effect if con is not passed.
        lazy_execution: If True, only re-executes if the node's definition has changed. Defaults to False. You should pass overwrite=True with lazy_execution as lazy_execution is intended precisely for iterative updates to a node definition. You must pass a connector (to cache results) for lazy_execution to work.
        n_threads: Max number of Node's to execute simultaneously when this node has multiple children.

    Returns:
        Table: The resulting table for this node. Also accessible through self.table after calling self.execute().
    """
    # Handle None tables
    if tables is None:
        tables = {}

    # Build dependency graph for all dependencies
    all_deps = self.dependencies
    nodes = {node.name: node for node in all_deps}
    nodes[self.name] = self  # Add self to the nodes

    # Build dependency and reverse graphs
    dependency_graph = self._build_dependency_graph(nodes)
    reverse_graph = self._build_reverse_graph(dependency_graph)

    # Track completion status and results
    completed = set()
    completion_lock = threading.Lock()
    worker_exceptions = []  # Track exceptions from worker threads
    stop_all_workers = (
        threading.Event()
    )  # Signal to stop all workers on first error

    # Track in-degree for scheduling
    in_degree = {}
    for node_name, dependencies in dependency_graph.items():
        in_degree[node_name] = len(dependencies)
    for node_name in nodes:
        if node_name not in in_degree:
            in_degree[node_name] = 0

    # Queue for nodes ready to execute
    ready_queue = queue.Queue()

    # Add nodes with no dependencies to ready queue
    for node_name, degree in in_degree.items():
        if degree == 0:
            ready_queue.put(node_name)

    def worker():
        """Worker function for thread pool"""
        while not stop_all_workers.is_set():
            try:
                node_name = ready_queue.get(timeout=1)
                # timeout forces to wait 1 second to avoid busy waiting
                if node_name is None:  # Sentinel value to stop worker
                    break
            except queue.Empty:
                continue

            try:
                logger.info(
                    f"Thread {threading.current_thread().name}: executing node '{node_name}'"
                )
                node = nodes[node_name]

                # Execute the node (without recursive child execution since we handle dependencies here)
                if lazy_execution:
                    if not overwrite:
                        raise ValueError(
                            "lazy_execution only works with overwrite=True."
                        )
                    if con is None:
                        raise ValueError(
                            "A DatabaseConnector is required for lazy execution."
                        )

                    if node._get_current_hash() != node._get_last_hash():
                        logger.info(f"Node '{node_name}': computing...")
                        table = node._execute(tables)
                        if (
                            table is not None
                        ):  # Only create table if _execute returns something
                            con.create_table(table, node_name, overwrite=overwrite)
                            table = con.get_dest_table(node_name)
                        node._update_current_hash()
                    else:
                        logger.info(
                            f"Node '{node_name}': unchanged, using cached result"
                        )
                        table = con.get_dest_table(node_name)
                else:
                    table = node._execute(tables)
                    if (
                        con and table is not None
                    ):  # Only create table if _execute returns something
                        con.create_table(table, node_name, overwrite=overwrite)
                        table = con.get_dest_table(node_name)

                node.table = table

                with completion_lock:
                    completed.add(node_name)

                    # Update in-degree for dependent nodes and add ready ones to queue
                    for dependent in reverse_graph.get(node_name, set()):
                        in_degree[dependent] -= 1
                        if in_degree[dependent] == 0:
                            # Check if all dependencies are completed
                            deps_completed = all(
                                dep in completed
                                for dep in dependency_graph.get(dependent, set())
                            )
                            if deps_completed:
                                ready_queue.put(dependent)

                logger.info(
                    f"Thread {threading.current_thread().name}: completed node '{node_name}'"
                )

            except Exception as e:
                logger.error(f"Error executing node '{node_name}': {str(e)}")
                with completion_lock:
                    # Store exception for main thread
                    worker_exceptions.append(e)
                    # Signal all workers to stop immediately and exit worker loop
                    stop_all_workers.set()
                    break
            finally:
                ready_queue.task_done()

    # Start worker threads
    threads = []
    for i in range(min(n_threads, len(nodes))):
        thread = threading.Thread(target=worker, name=f"PhenexWorker-{i}")
        thread.daemon = True
        thread.start()
        threads.append(thread)

    # Wait for all nodes to complete or for an error to occur
    while (
        len(completed) < len(nodes)
        and not worker_exceptions
        and not stop_all_workers.is_set()
    ):
        threading.Event().wait(0.1)  # Small delay to prevent busy waiting

    if not stop_all_workers.is_set():
        # Time to stop workers and cleanup
        stop_all_workers.set()

    # Check if any worker thread had an exception
    if worker_exceptions:
        # Signal workers to stop
        for _ in threads:
            ready_queue.put(None)
        # Wait for threads to finish
        for thread in threads:
            thread.join(timeout=1)
        # Re-raise the first exception
        raise worker_exceptions[0]

    # Signal workers to stop and wait for them
    for _ in threads:
        ready_queue.put(None)  # Sentinel value to stop workers

    for thread in threads:
        thread.join(timeout=1)

    logger.info(
        f"Node '{self.name}': completed multithreaded execution of {len(nodes)} nodes"
    )
    return self.table

visualize_dependencies()

Create a text visualization of the dependency graph for this node and its dependencies.

Returns:

Name	Type	Description
str	str	A text representation of the dependency graph

Source code in phenex/node.py

def visualize_dependencies(self) -> str:
    """
    Create a text visualization of the dependency graph for this node and its dependencies.

    Returns:
        str: A text representation of the dependency graph
    """
    lines = [f"Dependencies for Node '{self.name}':"]

    # Get all dependencies
    all_deps = self.dependencies
    nodes = {node.name: node for node in all_deps}
    nodes[self.name] = self  # Add self to the nodes

    # Build dependency graph
    dependency_graph = self._build_dependency_graph(nodes)

    for node_name in sorted(nodes.keys()):
        dependencies = dependency_graph.get(node_name, set())
        if dependencies:
            deps_str = ", ".join(sorted(dependencies))
            lines.append(f"  {node_name} depends on: {deps_str}")
        else:
            lines.append(f"  {node_name} (no dependencies)")

    return "\n".join(lines)