def getDatasetAsDataframe(user_id, knowage_address, dataset_name,
datastore_request):
address = "http://" + knowage_address + "/knowage/restful-services/2.0/datasets/" + dataset_name + "/data?offset=0&size=-1"
auth_token = security.buildAuthToken(user_id)
headers = {'Authorization': auth_token}
#rest request for dataset
payload = datastore_request
r = requests.post(address, headers=headers, data=payload)
#retrieve column names from metadata
names = r.json()["metaData"]["fields"]
column_names = []
column_types = {}
for x in names:
if type(x) is dict:
column_names.append(x['header'])
if x["type"] == "float":
column_types.update({x['name']: "float64"})
elif x["type"] == "float":
column_types.update({x['name']: "int64"})
#save data as dataframe
df = pd.DataFrame(r.json()["rows"])
#cast types
df = df.astype(column_types)
#drop first column (redundant)
df.drop(columns=['id'], inplace=True)
# assign column names
df.columns = column_names
return df
def getDatasetAsDataframe(widget):
address = widget.knowage_address + "/knowage/restful-services/2.0/datasets/" + widget.dataset_name + "/data?offset=0&size=-1"
auth_token = security.buildAuthToken(widget.user_id)
headers = {'Authorization': auth_token}
#rest request for dataset
payload = widget.datastore_request
r = requests.post(address, headers=headers, data=payload)
# retrieve column names from metadata
response_body = r.json()
if 'errors' in response_body:
raise KnowageRestServiceException(address)
col_names = response_body["metaData"]["fields"]
rows = response_body["rows"]
df = convertKnowageDatasetToDataframe(col_names, rows)
return df
def getDatasetAsDataframe(widget):
address = widget.knowage_address + "/knowage/restful-services/2.0/datasets/" + widget.dataset_name + "/data?offset=0&size=-1"
auth_token = security.buildAuthToken(widget.user_id)
headers = {'Authorization': auth_token}
#rest request for dataset
payload = widget.datastore_request
r = requests.post(address, headers=headers, data=payload)
#retrieve column names from metadata
names = r.json()["metaData"]["fields"]
#get cast types map
column_names, column_types = get_cast_types(names)
#save data as dataframe
df = pd.DataFrame(r.json()["rows"])
if not df.empty:
try:
#cast types
df = df.astype(column_types)
except Exception as e:
logging.warning("Could not cast dataframe types")
#drop first column (redundant)
df.drop(columns=['id'], inplace=True)
# assign column names
df.columns = column_names
return df