Pycaret throwing KeyError: 'UNSUPERVISED_DUMMY_TARGET' while running unsupervised anomaly detection on databricks notebook

Question

Error Screenshot I am trying to run setup function in pycaret==2.3.5 on python version 3.8.10 on databricks notebook and its showing the above error.

The same code is running fine on the local system but not on databricks. Can anyone help on this?

I tried changing the versions of pycaret and python but nothing worked.

Code used:

import pandas as pd
from pycaret.anomaly import *

df = pd.read_csv("global_master_data.csv")


df_copy = df.iloc[:,7:37].copy()
df_copy  = df_copy.drop(['Better For You Claim Text','First Selling OC','Manuf. Type','Private Label Brand - Desc'],axis = 1)


exp_name = setup(data = df_copy, silent = True)

Answer 1

I try to reproduce the same in my environment and got this output.

Install PyCaret library: Use this command pip install pycaret

I saved my dataset inside the file store. Using the below code I got the output.

import pandas as pd

df1 = pd.read_csv("/dbfs/FileStore/file_2.csv")

df_copy = df1.iloc[:,0:].copy()


df_copy  = df_copy.drop(['Name','subject'],axis = 1)

from pycaret.classification import *

exp_ano101 = setup(df_copy, normalize = True, session_id = 1, silent = True, html = False)

For more information refer this link1.