print "Error: cannot parse list of API (url, key) pairs. Each line in text file must contain the API url and API key for a single project separated by a space."
sys.exit()
#api_url_arch = api_pairs[0][0]
#api_key_arch = api_pairs[0][1]
api_url_ipss = api_pairs[1][0]
api_key_ipss = api_pairs[1][1]
api_url_psom = api_pairs[2][0]
api_key_psom = api_pairs[2][1]
# Get list of record IDs in IPSS. Exclude SickKids registry-only patients.
#record_ids = getRecordIDList(api_url_ipss, api_key_ipss)
#registry_data = exportRecords(api_url_ipss, api_key_ipss, record_id_list=record_ids, fields=["substud"], events=["acute_arm_1"])
#for row in registry_data:
# if (row["substud___8"] == "1"):
# record_ids.remove(row["ipssid"])
record_ids = getIPSSIDs(inc_registry_only=False, inc_unknown_stroke_type=False)
record_ids_post_2014 = getIPSSIDs(inc_registry_only=False,
inc_unknown_stroke_type=False,
inc_pre_2014=False)
record_ids_non_sk = getIPSSIDs(inc_registry_only=False,
inc_unknown_stroke_type=False,
inc_sk_patients=False)
record_ids_psom = getIPSSIDs(db="psom",
inc_registry_only=False,
inc_unknown_stroke_type=False)
# Load REDCap project (a PyCap object).
project = redcap.Project(api_url_ipss, api_key_ipss)
def_field = project.def_field
project_info = exportProjectInfo(api_url_ipss, api_key_ipss)
project_longitudinal = bool(project_info["is_longitudinal"])
def getPatientInfo(url_arch, url_ipss, key_arch, key_ipss):
## Get list of record IDs for each project. Exclude registry-only patients. Exclude patients with unknown stroke type.
# record_ids_arch = getRecordIDList(url_arch, key_arch)
# registry_arch = exportRecords(url_arch, key_arch, record_id_list=record_ids_arch, fields=["registry"], events=["acute_arm_1"])
# for row in registry_arch:
# if (row["registry"] == "1"):
# record_ids_arch.remove(row["pk_patient_id"])
# record_ids_ipss = getRecordIDList(url_ipss, key_ipss)
# registry_ipss = exportRecords(url_ipss, key_ipss, record_id_list=record_ids_ipss, fields=["substud"], events=["acute_arm_1"])
# for row in registry_ipss:
# if (row["substud___8"] == "1"):
# record_ids_ipss.remove(row["ipssid"])
# record_ids_ipss_only = [record_id for record_id in record_ids_ipss if not record_id in record_ids_arch]
# for record_id in record_ids_arch:
# if (not record_id in record_ids_ipss):
# print "Record with ID", record_id, "in Archive, but not in IPSS"
# Create one list of record ID which are non-registry and have known stroke type.
record_ids = getIPSSIDs(inc_registry_only=False,
inc_unknown_stroke_type=False)
## Create dict with patient information: {record_id: {dag:"...", enroll_date:"...", ...} }
patient_info = {}
for record_id in record_ids: # add item (another dict) for each patient in the Archive
patient_info[record_id] = {}
# patient_info[record_id]["in_arch"] = True
# if (record_id in record_ids_ipss):
# patient_info[record_id]["in_ipss"] = True # boolean describing presence of record in Archive
# else:
# patient_info[record_id]["in_ipss"] = False # boolean describing presence of record in IPSS
# for record_id in record_ids_ipss_only: # add item (another dict) for each patient in the IPSS that has not yet been added.
# patient_info[record_id] = {}
# patient_info[record_id]["in_arch"] = False
# patient_info[record_id]["in_ipss"] = True
## Get enrolment date for each record.
# Archive - Use 'dateofentry', then 'visit_date".
# print "Project : Archive"
dateofentry_arch = exportRecords(url_arch,
key_arch,
record_id_list=record_ids,
fields=["dateofentry"],
events=["acute_arm_1"],
validate=False)
# if (len(record_ids_arch) != len(dateofentry_arch)): # look for record id missing from exported data
# for record_id in record_ids_arch:
# id_in_data = False
# for row in dateofentry_arch:
# if (row["pk_patient_id"] == record_id):
# id_in_data = True
# break
# if (not id_in_data):
# print "Record with ID "+str(record_id)+" not found in exported data"
# num_missing = 0
for row in dateofentry_arch:
if (row["dateofentry"] == ""):
# num_missing += 1
pass
else:
if ("enroll_date" in patient_info[row["pk_patient_id"]]):
print "This record was counted twice: " + str(
row["pk_patient_id"])
continue
patient_info[row["pk_patient_id"]]["enroll_date"] = int(
row["dateofentry"][:4])
num_missing = len(
[id for id in record_ids if (not "enroll_date" in patient_info[id])])
# print "Field used : dateofentry"
# print "Number missing : ", num_missing
record_ids_leftover = [
id for id in record_ids if (not "enroll_date" in patient_info[id])
]
visit_date_leftover = exportRecords(url_arch,
key_arch,
record_id_list=record_ids_leftover,
fields=["visit_date"],
events=["acute_arm_1"],
validate=False)
# num_missing = 0
for row in visit_date_leftover:
if (row["visit_date"] == ""):
# num_missing += 1
pass
else:
if ("enroll_date" in patient_info[row["pk_patient_id"]]):
print "This record was counted twice: " + str(
row["pk_patient_id"])
continue
patient_info[row["pk_patient_id"]]["enroll_date"] = int(
row["visit_date"][:4])
num_missing = len(
[id for id in record_ids if (not "enroll_date" in patient_info[id])])
# print "Field used : visit_date"
# print "Number missing : ", num_missing
# IPSS - use 'dateentered' (works for all but 6 patients).
# print
# print "Project : IPSS"
record_ids_leftover = [
id for id in record_ids if (not "enroll_date" in patient_info[id])
]
dateentered_ipss = exportRecords(url_ipss,
key_ipss,
record_id_list=record_ids_leftover,
fields=["dateentered"],
events=["acute_arm_1"],
validate=False)
# if (len(record_ids_ipss_only) != len(dateentered_ipss)): # look for record id missing from exported data
# for record_id in record_ids_ipss_only:
# id_in_data = False
# for row in dateentered_ipss:
# if (row["ipssid"] == record_id):
# id_in_data = True
# break
# if (not id_in_data):
# print "Record with ID "+str(record_id)+" not found in exported data"
# num_missing = 0
for row in dateentered_ipss:
if (row["dateentered"] == ""):
# num_missing += 1
pass
else:
if ("enroll_date" in patient_info[row["ipssid"]]):
print "This record was counted twice: " + str(row["ipssid"])
continue
patient_info[row["ipssid"]]["enroll_date"] = int(
row["dateentered"][:4])
num_missing = len(
[id for id in record_ids if (not "enroll_date" in patient_info[id])])
# print "Field used : dateentered"
# print "Number missing : ", num_missing
enroll_dates = set()
for id, info in patient_info.iteritems():
if ('enroll_date' in info):
enroll_dates.add(info['enroll_date'])
if (not info['enroll_date'] in range(2003, 2020)):
print "Record enroll date outside [2003, 2019]:", id
else:
print "Record with no enrolment date:", id
# print "enroll_dates:", sorted(list(enroll_dates))
## Get DAG for each record:
dags_arch = exportRecords(url_arch,
key_arch,
record_id_list=record_ids,
fields=["pk_patient_id"],
validate=False)
dags_ipss = exportRecords(url_ipss,
key_ipss,
record_id_list=record_ids,
fields=["ipssid"],
validate=False)
for row in dags_arch:
record_id = row["pk_patient_id"]
dag = row["redcap_data_access_group"]
patient_info[record_id]["dag"] = dag
for row in dags_ipss:
record_id = row["ipssid"]
dag = row["redcap_data_access_group"]
if (not "dag" in patient_info[record_id]) or (
patient_info[record_id]["dag"]
== ""): # add DAG from IPSS if not added already
patient_info[record_id][
"dag"] = dag # overwriting DAG for records in Archive should not be a problem.
# for id in patient_info:
# if (not "dag" in patient_info[id]) or (patient_info[id]["dag"] == ""):
# print "Record with ID", id, "does not have a DAG assigned"
## Get stroke type for each patient. # Need to decide how we want to break this down further.
# stroke_type_arch = exportRecords(url_arch, key_arch, record_id_list=record_ids_arch, fields=["ais", "csvt", "pperi", "preart", "other_stroke", "age_at_event"], events=["acute_arm_1"])
stroke_type_ipss = exportRecords(url_ipss,
key_ipss,
record_id_list=record_ids,
fields=[
"chais", "chcsvt", "neoais",
"neocsvt", "ppis", "ppcsvt", "pvi",
"preart", "othcond"
],
events=["acute_arm_1"])
for record_id in patient_info:
patient_info[record_id]["stroke_type"] = {}
patient_info[record_id]["stroke_type"]["neo_ais"] = "2"
patient_info[record_id]["stroke_type"]["neo_csvt"] = "2"
patient_info[record_id]["stroke_type"]["child_ais"] = "2"
patient_info[record_id]["stroke_type"]["child_csvt"] = "2"
patient_info[record_id]["stroke_type"]["pp_ais"] = "2"
patient_info[record_id]["stroke_type"]["pp_csvt"] = "2"
patient_info[record_id]["stroke_type"]["pp_vi"] = "2"
patient_info[record_id]["stroke_type"]["art"] = "2"
patient_info[record_id]["stroke_type"]["other"] = "2"
for row in stroke_type_ipss: # 0 - no, 1 - yes, 2 - unknown
record_id = row["ipssid"]
# neonatal AIS
patient_info[record_id]["stroke_type"]["neo_ais"] = row["neoais___1"]
# neonatal CSVT
patient_info[record_id]["stroke_type"]["neo_csvt"] = row["neocsvt___1"]
# child AIS
patient_info[record_id]["stroke_type"]["child_ais"] = row["chais___1"]
# child CSVT
patient_info[record_id]["stroke_type"]["child_csvt"] = row[
"chcsvt___1"]
# presumed perinatal AIS
patient_info[record_id]["stroke_type"]["pp_ais"] = row["ppis___1"]
# presumed perinatal CSVT
patient_info[record_id]["stroke_type"]["pp_csvt"] = row["ppcsvt___1"]
# presumed perinatal VI
patient_info[record_id]["stroke_type"]["pp_vi"] = row["pvi___1"]
# arteriopathy
patient_info[record_id]["stroke_type"]["art"] = row["preart___1"]
# other
patient_info[record_id]["stroke_type"]["other"] = row["othcond___1"]
# Look for patients without an identified stroke type.
record_ids_with_unidentified_stroke_type = []
for id, record in patient_info.iteritems():
identified_type = False
for stroke_type, value in record["stroke_type"].iteritems():
if (value == "1"):
identified_type = True
break
if (not identified_type):
# print "Record with ID", id, "has an unidentified stroke type."
record_ids_with_unidentified_stroke_type.append(id)
# Check if stroke type can be identified in Archive instead.
# stroke_type_arch_leftover = exportRecords(url_arch, key_arch, record_id_list=record_ids_with_unidentified_stroke_type, fields=["ais", "csvt", "pperi", "preart", "other_stroke", "age_at_event"], events=["acute_arm_1"])
# for row in stroke_type_arch_leftover:
# print row["pk_patient_id"], row["ais"], row["csvt"], row["pperi"], row["preart"], row["other_stroke"]#, row["age_at_event"]
# stroke_type_found = False
# if (row["ais"] == "1") and (row["age_at_event"] == "0"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["neo_ais"] = "1"
# stroke_type_found = True
# if (row["csvt"] == "1") and (row["age_at_event"] == "0"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["neo_csvt"] = "1"
# stroke_type_found = True
# if (row["ais"] == "1") and (row["age_at_event"] == "1"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["child_ais"] = "1"
# stroke_type_found = True
# if (row["csvt"] == "1") and (row["age_at_event"] == "1"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["child_csvt"] = "1"
# stroke_type_found = True
# if (row["preart"] == "1"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["art"] = "1"
# stroke_type_found = True
# if (row["other_stroke"] == "1"):
# patient_info[row["pk_patient_id"]]["stroke_type"]["other"] = "1"
# stroke_type_found = True
# if stroke_type_found:
# record_ids_with_unidentified_stroke_type.remove(row["pk_patient_id"])
# Print some stats on the acquired patient information.
num_no_year = 0
num_no_dag = 0
for record_id, record in patient_info.iteritems():
if (record["dag"] == ""):
num_no_dag += 1
if (not "enroll_date" in record):
num_no_year += 1
print "Number of duplicated record IDs:", len(record_ids) - len(
set(record_ids))
print "Number of unique record IDs:", len(set(record_ids))
print "Number of record IDs in patient_info:", len(patient_info)
print "Number of records with no DAG:", num_no_dag
print "Number of records with no enrolment date:", num_no_year
print "Number of records with unidentified stroke type:", len(
record_ids_with_unidentified_stroke_type)
return patient_info
def buildProjects(config):
#### Read user's settings.yml file, which will be used to get API tokens and URLs.
api_settings = ApiSettings()
## Build a list of "projects" - dicts which store data and settings for the project.
projects = config["projects"]
## Verify the settings for each project.
for project in projects:
code_name = project["code_name"]
# Get args to pass to exportRecords.
if (not "exportRecords_args"
in project) or (project["exportRecords_args"] is None):
project["exportRecords_args"] = {}
# If use_getIPSSIDs is True, get list of record IDs to export.
if project["options"]["use_getIPSSIDs"]:
# If use_getIPSSIDs is True, but no options provided, raise warning.
if (not "getIPSSIDs_args"
in project) or (project["getIPSSIDs_args"] is None):
print "Warning: in project '" + code_name + "', 'use_getIPSSIDs' is True, but 'getIPSSIDs_args' not provided for project. Exporting all record IDs from project."
record_id_list = None
else:
getIPSSIDs_args = project["getIPSSIDs_args"]
record_id_list = getIPSSIDs(**getIPSSIDs_args)
# If exportRecords_args has an entry for record_id_list, but use_getIPSSIDs is True, raise warning.
if (project["options"]["use_getIPSSIDs"]) and (
"record_id_list" in project["exportRecords_args"]):
print "Warning: in project '" + code_name + "', the specified 'record_id_list' will be ignored, since 'use_getIPSSIDs' is True."
# Overwrite the record_id_list argument in exportRecords_args
project["exportRecords_args"]["record_id_list"] = record_id_list
## Get args to pass to exportRecords. If key does not exist, or it is not set to a value, set it to an empty dict (i.e.
exportRecords_args = project[
"exportRecords_args"] # has a value (possibly {}).
# Convert exportRecords_args arguments to strings as needed.
convert_to_strings = ["fields", "forms", "events", "record_id_list"]
for arg in convert_to_strings:
if arg in exportRecords_args.keys():
if (exportRecords_args[arg] == 'None'
): # these arguments could be lists or None
# Convert string 'None' to Python None.
exportRecords_args[arg] = None
else:
# Convert list to list of strings. Currently, list might contain integers etc.
new_list = [str(val) for val in exportRecords_args[arg]]
exportRecords_args[arg] = new_list
## Get API credentials for current project.
api_url, api_key, code_name = api_settings.getApiCredentials(
code_name=code_name)
project["api_url"] = api_url
project["api_key"] = api_key
## Export requested data for current project
data_csv = exportRecords(api_url,
api_key,
format="csv",
**exportRecords_args)
data_csv_file = StringIO(data_csv)
data_df = pandas.read_csv(data_csv_file,
dtype=unicode,
encoding='utf-8').fillna('')
project["chunks"] = [
data_df
] # this list of dataframes will be broken into pieces, each piece containing data to be placed in a different tab.
## Retrieve project settings and add them to the dict for the current project
pycap_project = redcap.Project(api_url, api_key)
def_field = pycap_project.def_field
project_info = exportProjectInfo(api_url, api_key)
longitudinal = bool(project_info["is_longitudinal"])
repeating = bool(project_info["has_repeating_instruments_or_events"])
events = getEvents(api_url, api_key, quiet=True)
metadata_raw = pycap_project.export_metadata()
form_event_mapping = exportFormEventMapping(pycap_project,
longitudinal)
repeating_forms_events = exportRepeatingFormsEvents(
api_url, api_key, repeating)
forms = exportFormsOrdered(api_url, api_key)
form_repetition_map = createFormRepetitionMap(longitudinal, repeating,
form_event_mapping,
repeating_forms_events,
forms)
metadata = parseMetadata(pycap_project.def_field,
project_info,
longitudinal,
repeating,
events,
metadata_raw,
form_event_mapping,
repeating_forms_events,
forms,
form_repetition_map,
write_branching_logic_function=False)
project["pycap_project"] = pycap_project
project["def_field"] = def_field
project["project_info"] = project_info
project["longitudinal"] = longitudinal
project["repeating"] = repeating
project["events"] = events
project["form_event_mapping"] = form_event_mapping
project["repeating_forms_events"] = repeating_forms_events
project["forms"] = forms
project["form_repetition_map"] = form_repetition_map
project["metadata"] = metadata
# Create dict which maps each form to a list of events containing that form.
if longitudinal:
form_to_events_dict = {}
for form_event_entry in form_event_mapping:
form = form_event_entry['form']
event = form_event_entry['unique_event_name']
if (not form in form_to_events_dict):
form_to_events_dict[form] = [event]
else:
form_to_events_dict[form].append(event)
else:
form_to_events_dict = None
project["form_to_events_dict"] = form_to_events_dict
## Build lists of variables which appear in the export data.
# columns which uniquely identify a row
primary_key = [def_field]
if project["longitudinal"]:
primary_key.append("redcap_event_name")
if project["repeating"]:
primary_key.append("redcap_repeat_instrument")
primary_key.append("redcap_repeat_instance")
project["primary_key"] = primary_key
primary_key_and_dag = primary_key
if ("redcap_data_access_group" in data_df.columns):
primary_key_and_dag.append("redcap_data_access_group")
project["primary_key_and_dag"] = primary_key_and_dag
# form_complete fields
form_complete_fields = [
field for field in data_df.columns
if ((field.endswith("_complete")) and (not field in metadata) and (
not field in primary_key) and (
not field == "redcap_data_access_group"))
]
project["form_complete_fields"] = form_complete_fields
# data fields
data_fields = [
field for field in data_df.columns
if ((not field in primary_key + form_complete_fields) and (
not field == "redcap_data_access_group"))
]
project["data_fields"] = data_fields
return projects
def getPatientInfo(url_arch, url_ipss, key_arch, key_ipss, enroll_date_min=2003, enroll_date_max=2020):
# Create one list of record ID which are non-registry and have known stroke type.
#record_ids = getIPSSIDs(ex_registry_only=True, ex_unknown_stroke_type=True, from_code_name="ipss_v3")
#print "DEBUG: CHANGE getIPSSIDs arguments back to IPSS V4."
record_ids = getIPSSIDs(ex_registry_only=True, ex_unknown_stroke_type=True)
## Create dict with patient information: {record_id: {dag:"...", enroll_date:"...", ...} }
patient_info = {}
for record_id in record_ids: # add item (another dict) for each patient in the Archive
patient_info[record_id] = {}
## Get enrolment date for each record.
# Archive - Use 'dateofentry', then 'visit_date".
dateofentry_arch = exportRecords(url_arch, key_arch, record_id_list=record_ids, fields=["dateofentry"], events=["acute_arm_1"], validate=False)
for row in dateofentry_arch:
if (row["dateofentry"] == ""):
pass
else:
if ("enroll_date" in patient_info[row["pk_patient_id"]]):
print "This record was counted twice: "+str(row["pk_patient_id"])
continue
patient_info[row["pk_patient_id"]]["enroll_date"] = int(row["dateofentry"][:4])
num_missing = len([id for id in record_ids if (not "enroll_date" in patient_info[id])])
record_ids_leftover = [id for id in record_ids if (not "enroll_date" in patient_info[id])]
visit_date_leftover = exportRecords(url_arch, key_arch, record_id_list=record_ids_leftover, fields=["visit_date"], events=["acute_arm_1"], validate=False)
for row in visit_date_leftover:
if (row["visit_date"] == ""):
pass
else:
if ("enroll_date" in patient_info[row["pk_patient_id"]]):
print "This record was counted twice: "+str(row["pk_patient_id"])
continue
patient_info[row["pk_patient_id"]]["enroll_date"] = int(row["visit_date"][:4])
num_missing = len([id for id in record_ids if (not "enroll_date" in patient_info[id])])
# IPSS - use 'dateentered' (works for all but 6 patients).
record_ids_leftover = [id for id in record_ids if (not "enroll_date" in patient_info[id])]
dateentered_ipss = exportRecords(url_ipss, key_ipss, record_id_list=record_ids_leftover, fields=["dateentered"], events=["acute_arm_1"], validate=False)
for row in dateentered_ipss:
if (row["dateentered"] == ""):
pass
else:
if ("enroll_date" in patient_info[row["ipssid"]]):
print "This record was counted twice: "+str(row["ipssid"])
continue
patient_info[row["ipssid"]]["enroll_date"] = int(row["dateentered"][:4])
num_missing = len([id for id in record_ids if (not "enroll_date" in patient_info[id])])
enroll_dates = set()
for id, info in patient_info.iteritems():
if ('enroll_date' in info):
enroll_dates.add(info['enroll_date'])
if (not info['enroll_date'] in range(enroll_date_min, enroll_date_max+1)):
print "Record enroll date outside ["+str(enroll_date_min)+", "+str(enroll_date_max)+"]:", id
else:
print "Record with no enrolment date:", id
## Get DAG for each record:
dags_arch = exportRecords(url_arch, key_arch, record_id_list=record_ids, fields=["pk_patient_id"], validate=False)
dags_ipss = exportRecords(url_ipss, key_ipss, record_id_list=record_ids, fields=["ipssid"], validate=False)
for row in dags_arch:
record_id = row["pk_patient_id"]
dag = row["redcap_data_access_group"]
patient_info[record_id]["dag"] = dag
for row in dags_ipss:
record_id = row["ipssid"]
dag = row["redcap_data_access_group"]
if (not "dag" in patient_info[record_id]) or (patient_info[record_id]["dag"] == ""): # add DAG from IPSS if not added already
patient_info[record_id]["dag"] = dag # overwriting DAG for records in Archive should not be a problem.
## Get stroke type for each patient. # Need to decide how we want to break this down further.
#stroke_type_ipss = exportRecords(url_ipss, key_ipss, record_id_list=record_ids, fields=["chais", "chcsvt", "neoais", "neocsvt", "ppis", "ppcsvt", "pvi", "preart", "othcond"], events=["acute_arm_1"])
stroke_type_ipss = exportRecords(url_ipss, key_ipss, record_id_list=record_ids, fields=["stroke_type"], events=["acute_arm_1"])
# Set stroke types to unknown initially.
for record_id in patient_info:
patient_info[record_id]["stroke_type"] = {}
patient_info[record_id]["stroke_type"]["neo_ais"] = "2"
patient_info[record_id]["stroke_type"]["neo_csvt"] = "2"
patient_info[record_id]["stroke_type"]["child_ais"] = "2"
patient_info[record_id]["stroke_type"]["child_csvt"] = "2"
patient_info[record_id]["stroke_type"]["pp_ais"] = "2"
patient_info[record_id]["stroke_type"]["pp_csvt"] = "2"
patient_info[record_id]["stroke_type"]["pp_vi"] = "2"
patient_info[record_id]["stroke_type"]["art"] = "2"
patient_info[record_id]["stroke_type"]["other"] = "2"
#'chais___1':'stroke_type___1',
#'chcsvt___1':'stroke_type___2',
#'neoais___1':'stroke_type___3',
#'neocsvt___1':'stroke_type___4',
#'ppis___1':'stroke_type___5',
#'ppcsvt___1':'stroke_type___6',
#'pvi___1':'stroke_type___7',
#'preart___1':'stroke_type___8',
#'othcond___1':'stroke_type___9'
for row in stroke_type_ipss: # 0 - no, 1 - yes, 2 - unknown
record_id = row["ipssid"]
# neonatal AIS
patient_info[record_id]["stroke_type"]["neo_ais"] = row["stroke_type___3"]
# neonatal CSVT
patient_info[record_id]["stroke_type"]["neo_csvt"] = row["stroke_type___4"]
# child AIS
patient_info[record_id]["stroke_type"]["child_ais"] = row["stroke_type___1"]
# child CSVT
patient_info[record_id]["stroke_type"]["child_csvt"] = row["stroke_type___2"]
# presumed perinatal AIS
patient_info[record_id]["stroke_type"]["pp_ais"] = row["stroke_type___5"]
# presumed perinatal CSVT
patient_info[record_id]["stroke_type"]["pp_csvt"] = row["stroke_type___6"]
# presumed perinatal VI
patient_info[record_id]["stroke_type"]["pp_vi"] = row["stroke_type___7"]
# arteriopathy
patient_info[record_id]["stroke_type"]["art"] = row["stroke_type___8"]
# other
patient_info[record_id]["stroke_type"]["other"] = row["stroke_type___9"]
# Look for patients without an identified stroke type.
record_ids_with_unidentified_stroke_type = []
for id, record in patient_info.iteritems():
identified_type = False
for stroke_type, value in record["stroke_type"].iteritems():
if (value == "1"):
identified_type = True
break
if (not identified_type):
record_ids_with_unidentified_stroke_type.append(id)
# Print some stats on the acquired patient information.
num_no_year = 0
num_no_dag = 0
for record_id, record in patient_info.iteritems():
if (record["dag"] == ""):
num_no_dag += 1
if (not "enroll_date" in record):
num_no_year += 1
print "Number of duplicated record IDs:", len(record_ids) - len(set(record_ids))
print "Number of unique record IDs:", len(set(record_ids))
print "Number of record IDs in patient_info:", len(patient_info)
print "Number of records with no DAG:", num_no_dag
print "Number of records with no enrolment date:", num_no_year
print "Number of records with unidentified stroke type:", len(record_ids_with_unidentified_stroke_type)
return patient_info