Goal

Create a project and upload your data in PredicSis.ai GUI using the Python SDK

Prerequisites

  • PredicSis.ai Python SDK (pip install predicsis; documentation)
  • Available data (can be downloaded here (≈95MB) )

In [1]:
# Load PredicSis.ai SDK
from predicsis import PredicSis

Create a project and upload data with one method


In [2]:
prj = PredicSis.create_project(
   'Outbound Mail Campaign', # naming the project
   {
       'name': 'Customers', # naming the table (=frame)
       'train': './Outbound Mail Campaign/Master_CENTRAL_TRAIN.csv', # choosing the train file
       'test': './Outbound Mail Campaign/Master_CENTRAL_TEST.csv' # choosing the test file
   },
   peripherals=[ # adding peripheral tables
       {
           'name': 'Email', # naming the peripheral table (=frame)
           'train': './Outbound Mail Campaign/Email_TRAIN.csv', # choosing the train file
           'test': './Outbound Mail Campaign/Email_TEST.csv', # choosing the test file
       },
       {
           'name': 'Orders', # naming another peripheral table (=frame)
           'train': './Outbound Mail Campaign/Orders_TRAIN.csv', # choosing the train file
           'test': './Outbound Mail Campaign/Orders_TEST.csv', # choosing the test file
       },
       {
           'name': 'VisitedPages', # naming another peripheral table (=frame)
           'train': './Outbound Mail Campaign/VisitedPages_TRAIN.csv', # choosing the train file
           'test': './Outbound Mail Campaign/VisitedPages_TEST.csv', # choosing the test file
       }
   ],
   key='KEY', # choosing the identifier, also used as the join key
   target='LABEL', # name of the outcome feature
   main_modality='1', # modality of the outcome feature on which you want the reports and the scores
   separator='|'# field separator in all files
 )

In [ ]: