re-build a dataset from dataiku webapp

rakesh99
Level 1
re-build a dataset from dataiku webapp

I trying to create project ,where I can recalculate my data based the user input . below is my flow in dataiku project 



Data - this data set ,i have uploaded through csv file 



webapp_input - This dataset ,I am loading  from webapp (created in html,js and python backend )



Output- this data set I am creating using python recipe



Problem statement :



whenever ,I am adding more data in webapp_input through dataiku webapp .I have to run the python receipt manually  to re-build the "output"dataset  . 



But I am trying to trigger it through webapp itself .As soon as will update the data in UI .it should store in webapp_input dataset (which is working fine ) and re-build the "output" dataset .



 



I have tried following python scripts for this .



import pandas as pd



import numpy as np



import dataiku



import json

import dataikuapi



client = dataiku.api_client()



def jobrun():



    project = client.get_project('Testing')

    definition = {

        "type" : "NON_RECURSIVE_FORCED_BUILD",

        "outputs" : [{

                      "id" : "output",

                       "partition" : "NP"

                                }]

                    }

    job = project.start_job(definition)

    state = ''

    while state != 'DONE' and state != 'FAILED' and state != 'ABORTED':

       time.sleep(1)

        state = job.get_status()['baseStatus']['state']



Even though the dataset is present in the Project, itโ€™s giving us below error message.



DataikuException: java.lang.IllegalArgumentException: Dataset not found or not buildable: Testing.output 



 



I have tried second option :



from dataiku.scenario import Scenario

client = dataiku.api_client()



scenario.build_dataset("output ") 



 



Non of the options are working .would you be to help me with solving this issue .or any other approach for my requirement .



Thanks



 

0 Kudos
3 Replies
Clรฉment_Stenac
Hi,

Are you sure that your project key is "Testing" ? The project key is not the name that displays on the home page, but the portion in the URL. It is usually all-caps
0 Kudos
rakesh99
Level 1
Author
Hi Stenac,

Thanks for your quick response .
"Testing " is my project name and I am writing this script within the dataiku webapp .Do I still need to get the project key for that .

Can you please suggest ,where I should put my project key

import pandas as pd

import numpy as np

import dataiku

import json
import dataikuapi
import dataiku, os.path
from flask import Flask

handle = dataiku.Folder("my")

path = handle.get_path()




def jobrun():
client = dataiku.api_client()
project = client.get_project('Testing')
definition = {
"type" : "NON_RECURSIVE_FORCED_BUILD",
"outputs" : [{
"id" : "output",
"partition" : "NP"
}]
}
job = project.start_job(definition)
#state = ''
#while state != 'DONE' and state != 'FAILED' and state != 'ABORTED':
#time.sleep(1)
#state = job.get_status()['baseStatus']['state']


jobrun()













I was referring below link :

https://doc.dataiku.com/dss/latest/publicapi/client-python/jobs.html
0 Kudos
rakesh99
Level 1
Author
Thanks ... Stenac

I am able to run it now .
0 Kudos

Labels

?
Labels (2)
A banner prompting to get Dataiku