def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"period.start",
"period.end",
],
code_columns=[
*expand_args.get("code_columns", []),
"class",
"priority",
"participant",
"length",
"hospitalization",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("period"),
Frame.codeable_like_column_expander("reason"),
Frame.codeable_like_column_expander("location"),
Frame.codeable_like_column_expander("serviceProvider"),
],
)
def transform_results(data_frame: pd.DataFrame, params={}, **expand_args):
def expand_id(id_column: pd.Series):
return pd.concat(
[
id_column,
id_column.str.split(":", expand=True).rename(columns={
0: "variant_set_id",
2: "gene"
})[["variant_set_id", "gene"]],
],
axis=1,
)
args = {
**expand_args,
"custom_columns": [
*expand_args.get("custom_columns", []),
*[
Frame.codeable_like_column_expander(k) for k in [
"clinvar",
"cosmic",
"vcf",
"ensemblCanon",
"dbnsfp",
]
],
("id", expand_id),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"authoredOn",
"dispenseRequest.validityPeriod.start",
"dispenseRequest.validityPeriod.end",
],
code_columns=[
*expand_args.get("code_columns", []),
"medicationCodeableConcept",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("context"),
Frame.codeable_like_column_expander("note"),
(
"dispenseRequest",
lambda r: pd.json_normalize(r).add_prefix(
"dispenseRequest."),
),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
custom_columns=[
*expand_args.get("custom_columns", []),
("name", expand_name_column),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
code_columns=[*expand_args.get("code_columns", []), "type"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[*expand_args.get("date_columns", []), "dateTime"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("sourceReference"),
Frame.codeable_like_column_expander("actor"),
Frame.codeable_like_column_expander("patient"),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[*expand_args.get("date_columns", []), "date"],
code_columns=[*expand_args.get("code_columns", []), "vaccineCode"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("patient"),
Frame.codeable_like_column_expander("encounter"),
],
)
def transform_results(data_frame: pd.DataFrame, **expand_args):
args = {
**expand_args,
"date_columns":
[*expand_args.get("date_columns", []), "startDate"],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
code_columns=[
*expand_args.get("code_columns", []),
"procedureCode",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("patient"),
Frame.codeable_like_column_expander("context"),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"recorded",
"signature.when",
],
code_columns=[
*expand_args.get("code_columns", []),
"agent",
"signature",
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
code_columns=[*expand_args.get("code_columns", []), "type"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("context"),
(
"requester",
lambda r: pd.json_normalize(r).add_prefix("requester."),
),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
code_columns=[
*expand_args.get("code_columns", []),
"medicationCodeableConcept",
"quantity",
"dosageInstruction",
"daysSupply",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
],
)
def transform_results(data_frame: pd.DataFrame, **expand_args):
args = {
**expand_args,
"code_columns": [
*expand_args.get("code_columns", []),
"specimen",
"repository",
],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("patient"),
Frame.codeable_like_column_expander("referenceSeq"),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"effectiveDateTime",
],
code_columns=[
*expand_args.get("code_columns", []),
"medicationCodeableConcept",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
code_columns=[*expand_args.get("code_columns", []), "type"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
# TODO: Properly parse content column
#
# Example:
# [{'attachment': {'contentType': 'application/gzip',
# 'url': 'https://api.us.lifeomic.com/v1/files/<uuid>',
# 'size': 182539,
# 'title': 'helix-source-files/normalized/<filename>.vcf.gz'}}]
],
)
def transform_results(data_frame: pd.DataFrame, **expand_args):
args = {
**expand_args,
"code_columns": [
*expand_args.get("code_columns", []),
"contained",
"maritalStatus",
],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("managingOrganization"),
("address", expand_address_column),
("name", expand_name_column),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"occurrenceDateTime",
],
code_columns=[*expand_args.get("code_columns", []), "bodySite"],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
(
"content",
lambda r: pd.json_normalize(r).add_prefix("content."),
),
],
)
def test_frame_expand_date_out_of_range():
original = pd.DataFrame(
[
{"effectiveDateTime": "2020-09-15 12:31:00-0500", "id": "obs1"},
{"effectiveDateTime": "0217-05-04 12:31:00-0500", "id": "obs2"},
]
)
expanded = Frame.expand(original)
assert expanded.at[0, "effectiveDateTime.tz"] == -5.0
assert expanded.at[0, "effectiveDateTime.local"] == pd.Timestamp(
"2020-09-15 12:31:00", tz="utc"
)
assert math.isnan(expanded.at[1, "effectiveDateTime.tz"])
assert pd.isna(expanded.at[1, "effectiveDateTime.local"])
def transform_results(data_frame: pd.DataFrame, **expand_args):
args = {
**expand_args,
"date_columns": [
*expand_args.get("date_columns", []),
"performedPeriod.start",
"performedPeriod.end",
],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("performedPeriod"),
Frame.codeable_like_column_expander("context"),
Frame.codeable_like_column_expander("managingOrganization"),
],
}
return Frame.expand(data_frame, **args)
def transform_results(data_frame: pd.DataFrame, **expand_args):
args = {
**expand_args,
"code_columns": [
*expand_args.get("code_columns", []),
"component",
"interpretation",
],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("related"),
Frame.codeable_like_column_expander("performer"),
Frame.codeable_like_column_expander("context"),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"occurrencePeriod.start",
"occurrencePeriod.end",
"occurrenceDateTime",
"authoredOn",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("context"),
Frame.codeable_like_column_expander("occurrencePeriod"),
Frame.codeable_like_column_expander("note"),
],
)
def test_local_and_timezone_split():
original = pd.DataFrame(
[
{"effectiveDateTime": "2020-08-08 11:00:00+0300", "id": "obs1"},
{"effectiveDateTime": "2020-08-09 10:00:00-0400", "id": "obs2"},
]
)
expanded = Frame.expand(original)
assert expanded.at[0, "effectiveDateTime.tz"] == 3.0
assert expanded.at[0, "effectiveDateTime.local"] == pd.Timestamp(
"2020-08-08 11:00:00", tz="utc"
)
assert expanded.at[1, "effectiveDateTime.tz"] == -4.0
assert expanded.at[1, "effectiveDateTime.local"] == pd.Timestamp(
"2020-08-09 10:00:00", tz="utc"
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"effectiveDateTime",
],
code_columns=[
*expand_args.get("code_columns", []),
"type",
"subtype",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("agent"),
Frame.codeable_like_column_expander("source"),
Frame.codeable_like_column_expander("entity"),
],
)
def transform_results(
data_frame: pd.DataFrame, params: dict, **expand_args
):
args = {
**expand_args,
"code_columns": [
*expand_args.get("code_columns", []),
"bodySite",
"patient",
],
"custom_columns": [
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("sourceFile"),
],
}
df = Frame.expand(data_frame, **args)
if "sets" in df.columns:
df = (
pd.concat(
data_frame.apply(
lambda x: pd.DataFrame(
[{"index": x.name, **s} for s in x.sets]
),
axis=1,
).values
)
.join(df.drop(["sets"], axis=1), on="index", rsuffix=".test")
.drop(["index"], axis=1)
.reset_index(drop=True)
)
test_type = params.get("type", None)
if test_type and len(df) > 0:
# TODO: Remove when API fixed
# NOTE: The API does not filter the returned sets because it is a
# nested structure. Since it's not a boatload of information, we opt
# to filter client-side for now.
return df[df.setType == test_type].reset_index(drop=True)
return df
def transform_results(data_frame: pd.DataFrame, params={}, **expand_args):
def expand_id(id_column: pd.Series):
return pd.concat(
[
id_column,
id_column.str.split(":", expand=True).rename(
columns={0: "variant_set_id"})["variant_set_id"],
],
axis=1,
)
args = {
**expand_args,
"custom_columns": [
*expand_args.get("custom_columns", []),
("id", expand_id),
],
}
return Frame.expand(data_frame, **args)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"effectivePeriod.start",
"effectivePeriod.end",
],
code_columns=[
*expand_args.get("code_columns", []),
"medicationCodeableConcept",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("context"),
Frame.codeable_like_column_expander("prescription"),
Frame.codeable_like_column_expander("dosage"),
Frame.codeable_like_column_expander("effectivePeriod"),
],
)
def transform_results(df: pd.DataFrame, **expand_args):
return Frame.expand(
df,
date_columns=[
*expand_args.get("date_columns", []),
"onsetDateTime",
"assertedDate",
"onsetPeriod.start",
"onsetPeriod.end",
],
code_columns=[
*expand_args.get("code_columns", []),
"bodySite",
"stage",
],
custom_columns=[
*expand_args.get("custom_columns", []),
Frame.codeable_like_column_expander("subject"),
Frame.codeable_like_column_expander("onsetPeriod"),
Frame.codeable_like_column_expander("context"),
],
)