Make a report from the JSON report of a CDISC CORE run.
from_core_report(
report,
package=None,
agency=None,
)
CDISC CORE can be run on its own (as core validate ... -of JSON), and its JSON report can be kept for later. The from_core_report() class method reads such a report into a ConformanceReport, which gives it the same methods as the report of validate_cdisc_submission(). The JSON report has five sections: Conformance_Details (the standard, its version and the version of CORE), Dataset_Details, Issue_Summary (the number of issues for each rule and dataset), Issue_Details (one entry for each failing record) and Rules_Report (the status of each rule). A file written by to_json() from a CORE report has the same sections, so it can be read back with this method.
Parameters
report: dict | ParsedCoreReport
-
The CORE report, as a dict (such as from json.load()) or as the already parsed report of an earlier ConformanceReport (its core attribute).
package: SubmissionPackage | None = None
-
The SubmissionPackage that CORE checked. The default is None.
agency: str | None = None
-
The regulatory agency, such as
"FDA", which is only used as a label in the report. The default is None.
Returns
ConformanceReport
-
A report of CORE’s results, whose
is_core is True.
Examples
Let’s write a short report by hand in CORE’s format, for three rules on a DM and an AE dataset. One rule found an AE record whose subject isn’t in DM, one rule passed and one was skipped:
import pointblank as pb
core_json = {
"Conformance_Details": {
"CORE_Engine_Version": "0.16.0",
"Standard": "SDTMIG",
"Version": "V3.4",
},
"Dataset_Details": [
{"filename": "dm.xpt", "label": "Demographics", "length": 3},
{"filename": "ae.xpt", "label": "Adverse Events", "length": 2},
],
"Issue_Summary": [
{
"dataset": "ae.xpt",
"core_id": "CORE-000201",
"message": "USUBJID is not found in DM.USUBJID",
"issues": 1,
}
],
"Issue_Details": [
{
"core_id": "CORE-000201",
"message": "USUBJID is not found in DM.USUBJID",
"dataset": "ae.xpt",
"USUBJID": "STUDY01-009",
"row": 2,
"SEQ": 1,
"variables": ["USUBJID"],
"values": ["STUDY01-009"],
}
],
"Rules_Report": [
{"core_id": "CORE-000201", "status": "ISSUE REPORTED"},
{"core_id": "CORE-000351", "status": "SUCCESS"},
{"core_id": "CORE-000087", "status": "SKIPPED"},
],
}
report = pb.ConformanceReport.from_core_report(core_json, agency="FDA")
print(report)
ConformanceReport (CORE)
Agency: FDA
SDTMIG V3.4 — CORE 0.16.0
3 rules (ISSUE REPORTED=1, SKIPPED=1, SUCCESS=1)
1 issues — FAIL
The rule statuses are CORE’s own. A real report would usually be read from a file, with json.load(). The failing record is in findings():
[CoreFinding(rule_id='CORE-000201', message='USUBJID is not found in DM.USUBJID', dataset='ae.xpt', executability=None, usubjid='STUDY01-009', row=2, seq=1, variables=['USUBJID'], values=['STUDY01-009'])]
CORE gives the row of a record as a 1-based number, so row=2 is the second record of the AE dataset.