'DSSClient' object has no attribute 'predict_record'

Options
shajay12
shajay12 Registered Posts: 14 ✭✭✭
edited July 16 in Using Dataiku

Hi Everyone,

Once please help on this. I get this error for DSS client. Thanks in Advance.

2_deploy_dev/test_dev.py:37: in <module> prediction = client_dep.predict_record(api_endpoint_id, record_to_predict) E AttributeError: 'DSSClient' object has no attribute 'predict_record'

at this line

prediction = client_dep.predict_record(api_endpoint_id, record_to_predict)

Code here:

import dataikuapi
import sys

host = sys.argv[2]
apiKey = sys.argv[3]
api_service_id = sys.argv[4]
api_endpoint_id = sys.argv[5]
api_dev_infra_id = sys.argv[6]

client_dep = dataikuapi.DSSClient(host,apiKey)
client_dep._session.verify = False

record_to_predict = {

    "XXXXXXXCOUNTRY": "XXX",

    "XXXXXXXXXXXX": XXXX,

    "FXXXXXXXXXX": "XXX",

    "XXXXXXXXXXXXXXXXXX": "XXXXXXXXXXXXXK",

    "XXXXXXXXXXXX": XXXXX,

    "XXXXXXXXXXAG": "X",

    "XXXXXXXXXXXXr": "1XXXXXXX",

    "cXXXXXXX": "XX",

    "LXXXXXXde": "XXXXXM",

    "HXXXXXX": "9XXXXXX"

}
prediction = client_dep.predict_record(api_endpoint_id, record_to_predict)
assert prediction['result']['prediction'] == '1', "Prediction should be 1 but is {}".format(prediction['result']['prediction'])

Answers

  • louisplt
    louisplt Dataiker Posts: 21 Dataiker
    edited 3:51PM
    Options

    Hello @shajay12
    ,

    As your are requesting an API Node, you need to use an APINodeClient instead of a DSSClient:

    client = dataikuapi.APINodeClient(uri, service_id, api_key)

    See https://doc.dataiku.com/dss/latest/apinode/api/user-api.html for more details.

    Hope this helps

  • shajay12
    shajay12 Registered Posts: 14 ✭✭✭
    Options

    Hi Louisplt,

    I used the APINodeClient, but it gives this error.

    2_deploy_dev/test_dev.py:39: in <module> prediction = client_dep.predict_record(api_endpoint_id, record_to_predict) /usr/local/lib/python3.6/dist-packages/dataikuapi/apinode_client.py:54: in predict_record return self._perform_json("POST", "%s/predict" % endpoint_id, body = obj) /usr/local/lib/python3.6/dist-packages/dataikuapi/base_client.py:41: in _perform_json return self._perform_http(method, path, params, body, False).json() /usr/local/lib/python3.6/dist-packages/dataikuapi/base_client.py:27: in _perform_http auth=auth, stream = stream) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:522: in request resp = self.send(prep, **send_kwargs) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:636: in send adapter = self.get_adapter(url=request.url) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:727: in get_adapter raise InvalidSchema("No connection adapters were found for '%s'" % url) E requests.exceptions.InvalidSchema: No connection adapters were found for '-o/public/api/v1/--host=https://apideployer-01.design.dataiku.XXXX.com:11000//--api=o1pz4GQUPGD1vE8YqoOq8wuUd0824xzs/predict'

  • louisplt
    louisplt Dataiker Posts: 21 Dataiker
    Options

    Hello @shajay12

    Could you please share the code you used to call APINodeClient.predict_record?

  • kurt
    kurt Registered Posts: 1 ✭✭✭
    edited 3:51PM
    Options

    @louisplt
    wrote:

    Hello @shajay12
    ,

    As your are requesting an API Node, you need to use an APINodeClient instead of a DSSClient:

    client = dataikuapi.APINodeClient(uri, service_id, api_key)

    See https://doc.dataiku.com/dss/latest/apinode/api/user-api. for more details.

    Hope this helps

    Ein Metall-Gartenzaun ist eine robuste und pflegeleichte Lösung, um Ihren Garten zu begrenzen. Die Montage eines Zauns aus Metall von den Zaun-Experten bietet eine sinnvolle Möglichkeit, Kinder und Haustiere im Garten zu sichern. Die passenden Pflanzkübel kaufen sie für Ihren Garten und Terrasse bei https://www.pflanzkuebel7.de/.

    Definitiv empfhelenswerte Experten für Zäune und Pflanzkübel!


    Thanks. This worked for me!

  • shajay12
    shajay12 Registered Posts: 14 ✭✭✭
    edited 3:51PM
    Options

    Hi louisplt, Thanks for the reply and help.

    here is the code.

    import dataikuapi
    
    host = sys.argv[2]
    apiKey = sys.argv[3]
    api_service_id = sys.argv[4]
    api_endpoint_id = sys.argv[5]
    api_dev_infra_id = sys.argv[6]
    
    client_dep = dataikuapi.APINodeClient(host,api_service_id,apiKey)
    client_dep._session.verify = False
    
    record_to_predict = {
        "XXXX-XXXX-XXX": "XXX",
        "XXXXXXXXXXXX": 1XXXXX,
        "XXXXXXXXXXXXXXXXYPE": "XXXXXX",
        "XXXXXXXXXX-1": "XXXXXXXXXK",
        "XXXXXXXXXXXXXXXE": XXXXX,
        "XXXXXXXXXXXXXXG": "X",
        "XXXXXXXXXXX": "xxxxxx",
        "xxxxxxxxxxxx": "xxx",
        "xxxxxxxxxxxx": "xxxxxxxx",
        "xxxxxxxxxxxxxxx": "xxxxxxxxxxxxx"
    }
    prediction = client_dep.predict_record(api_endpoint_id, record_to_predict)
    assert prediction['result']
    ['prediction'] == '1', "Prediction should be 1 but is {}".format(prediction['result']['prediction'])

    also here is the YAML code that is calling the above Python script in Azure pipeline.

    python3 -m pytest 2_deploy_dev/test_dev.py -o junit_family=xunit1 --host='${{ parameters.DEPLOYER_URL }}' --api='${{ parameters.DEPLOYER_API_KEY }}' --api_service_id='${{ parameters.API_SERVICE_ID }}' --api_endpoint_id='${{ parameters.API_ENDPOINT_ID }}' --api_dev_infra_id='${{ parameters.API_DEV_INFRA_ID }}' --junitxml=reports/DEV_TEST.xml

    error message.

    2_deploy_dev/test_dev.py:39: in <module> prediction = client_dep.predict_record(api_endpoint_id, record_to_predict) /usr/local/lib/python3.6/dist-packages/dataikuapi/apinode_client.py:54: in predict_record return self._perform_json("POST", "%s/predict" % endpoint_id, body = obj) /usr/local/lib/python3.6/dist-packages/dataikuapi/base_client.py:41: in _perform_json return self._perform_http(method, path, params, body, False).json() /usr/local/lib/python3.6/dist-packages/dataikuapi/base_client.py:27: in _perform_http auth=auth, stream = stream) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:522: in request resp = self.send(prep, **send_kwargs) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:636: in send adapter = self.get_adapter(url=request.url) /usr/local/lib/python3.6/dist-packages/requests/sessions.py:727: in get_adapter raise InvalidSchema("No connection adapters were found for '%s'" % url) E requests.exceptions.InvalidSchema: No connection adapters were found for '-o/public/api/v1/--host=https://apideployer-01.design.dataiku.fedex.com:11000//--api=o1pz4GQUPGD1vE8YqoOq8wuUd0824xzs/predict'

  • shajay12
    shajay12 Registered Posts: 14 ✭✭✭
    edited 3:51PM
    Options

    Hi Kurt,

    Thanks for your reply and help. Once check the code that I have replied to in the previous post.

    Could you please tell me the value of these variables 'apinodes' and 'url' in this link

    Though I took code from this link https://knowledge.dataiku.com/latest/kb/o16n/ci-cd/cicd-with-api-deployer.html

    in the above link, I was not able to find the value of 'apinodes' and 'url' variables, so I had made changes to the code. But the changed code also did not work.

    Below is the code in the above link

    def build_apinode_client(params):
        client_design = dataikuapi.DSSClient(params["host"], params["api"])
        api_deployer = client_design.get_apideployer()
        api_url = api_deployer.get_infra(params["api_dev_infra_id"]).get_settings().get_raw()['apiNodes'][0]['url']
        return dataikuapi.APINodeClient(api_url, params["api_service_id"])

  • shajay12
    shajay12 Registered Posts: 14 ✭✭✭
    edited 3:51PM
    Options

    I have used APINodeClient code also,

    but it gives the same error.

    client_dep = dataikuapi.APINodeClient(host,api_service_id,apiKey)
    client_dep._session.verify = False
    
    record_to_predict = {
        "XXXXXXXCOUNTRY": "XXX",
        "XXXXXXXXXXXX": XXXX,
        "FXXXXXXXXXX": "XXX",
        "XXXXXXXXXXXXXXXXXX": "XXXXXXXXXXXXXK",
        "XXXXXXXXXXXX": XXXXX,
        "XXXXXXXXXXAG": "X",
        "XXXXXXXXXXXXr": "1XXXXXXX",
        "cXXXXXXX": "XX",
        "LXXXXXXde": "XXXXXM",
        "HXXXXXX": "9XXXXXX"
    }
    prediction = client_dep.predict_record(api_endpoint_id, record_to_predict)
    assert prediction['result']['prediction'] == '1', "Prediction should be 1 but is {}".format(prediction['result']['prediction'])

Setup Info
    Tags
      Help me…