Spark

TLDR

import edsnlp

stream = edsnlp.data.from_spark(df, converter="omop")
stream = stream.map_pipeline(nlp)
res = stream.to_spark(converter="omop")
# or equivalently
edsnlp.data.to_spark(stream, converter="omop")

We provide methods to read and write documents (raw or annotated) from and to Spark DataFrames.

As an example, imagine that we have the following OMOP dataframe (we'll name it note_df)

note_id	note_text	note_datetime
0	Le patient est admis pour une pneumopathie...	2021-10-23

Reading from a Spark Dataframe[source]

The SparkReader (or edsnlp.data.from_spark) reads a pyspark (or koalas) DataFrame and yields documents. At the moment, only entities and span attributes are loaded.

Example

import edsnlp

nlp = edsnlp.blank("eds")
nlp.add_pipe(...)
doc_iterator = edsnlp.data.from_spark(note_df, converter="omop")
annotated_docs = nlp.pipe(doc_iterator)

Generator vs list

edsnlp.data.from_spark returns a Stream To iterate over the documents multiple times efficiently or to access them by index, you must convert it to a list

docs = list(edsnlp.data.from_spark(note_df, converter="omop"))

Parameters

PARAMETER	DESCRIPTION
`data`	The DataFrame to read.
`shuffle`	Whether to shuffle the data. If "dataset", the whole dataset will be shuffled before starting iterating on it (at the start of every epoch if looping). TYPE: `Literal['dataset', False]` DEFAULT: `False`
`seed`	The seed to use for shuffling. TYPE: `Optional[int]` DEFAULT: `None`
`loop`	Whether to loop over the data indefinitely. TYPE: `bool` DEFAULT: `False`
`converter`	Converters to use to convert the rows of the DataFrame to Doc objects. These are documented on the Converters page. TYPE: `Optional[AsList[Union[str, Callable]]]` DEFAULT: `None`
`kwargs`	Additional keyword arguments to pass to the converter. These are documented on the Converters page. DEFAULT: `{}`

RETURNS	DESCRIPTION
`Stream`

Writing to a Spark DataFrame[source]

edsnlp.data.to_spark converts a list of documents into a Spark DataFrame, usually one row per document, unless the converter returns a list in which case each entry of the resulting list will be stored in its own row.

Example

import edsnlp, edsnlp.pipes as eds

nlp = edsnlp.blank("eds")
nlp.add_pipe(eds.covid())

note_df = sql('''
    select note_id, note_text from note
    where note_text is not null
    limit 500
''')

docs = edsnlp.data.from_spark(note_df, converter="omop")

docs = nlp.pipe(docs)

res = edsnlp.data.to_spark(docs, converter="omop")

res.show()

Mac OS X

If you are using Mac OS X, you may need to set the following environment variable (see this thread) to run pyspark:

import os
os.environ["OBJC_DISABLE_INITIALIZE_FORK_SAFETY"] = "YES"

Parameters

PARAMETER	DESCRIPTION
`data`	The data to write (either a list of documents or a Stream). TYPE: `Union[Any, Stream]`
`dtypes`	The schema to use for the DataFrame. TYPE: `Any` DEFAULT: `None`
`show_dtypes`	Whether to print the inferred schema (only if `dtypes` is None). TYPE: `bool` DEFAULT: `True`
`execute`	Whether to execute the writing operation immediately or to return a stream TYPE: `bool` DEFAULT: `True`
`converter`	Converter to use to convert the documents to dictionary objects before storing them in the dataframe. These are documented on the Converters page. TYPE: `Optional[Union[str, Callable]]` DEFAULT: `None`
`kwargs`	Additional keyword arguments to pass to the converter. These are documented on the Converters page. DEFAULT: `{}`

Importing entities from a Spark DataFrame

If you have a dataframe with entities (e.g., note_nlp in OMOP), you must join it with the dataframe containing the raw text (e.g., note in OMOP) to obtain a single dataframe with the entities next to the raw text. For instance, the second note_nlp dataframe that we will name note_nlp.

note_nlp_id	note_id	start_char	end_char	note_nlp_source_value	lexical_variant
0	0	46	57	disease	coronavirus
1	0	77	88	drug	paracétamol

import pyspark.sql.functions as F

df = note_df.join(
    note_nlp_df
    .groupBy("note_id")
    .agg(
        F.collect_list(
            F.struct(
                F.col("note_nlp_id"),
                F.col("start_char"),
                F.col("end_char"),
                F.col("note_nlp_source_value")
            )
        ).alias("entities")
    ), "note_id", "left")