Installation and usage
set up python environments
#connect python environment with tableau
pip install tabpy
#connect jupyter environment with tableau
pip install tabpy_client
run tabpy
run following command: tabpy will start a python local server on port 9004. after connecting tableau with tabpy server, you can implement python script
deployment ML model to tableau sample
an interactive dashboard with optimal variables to predict the probability of fraud and late delivery
#after connecting tabpy server to Tableau, you can deploy ML model to tableau
import tabpy_client
from tabpy.tabpy_tools.client import Client
client = tabpy_client.Client('http://localhost:9004/')
def fraud_predictor5( _arg1, _arg2,_arg3):
import pandas as pd
row = {'shipping': _arg1,
'shipping scheduled': _arg2,
'country_str':_arg3}
#Convert it into a dataframe
test_data = pd.DataFrame(data = row,index=[0])
from sklearn import preprocessing
le = preprocessing.LabelEncoder()
test_data['country_str'] = le.fit_transform(test_data['country_str'])
#Predict the Fraud
predprob_survival = random_forest.predict_proba(test_data)
#Return only the probability
return [probability[1] for probability in predprob_survival]
def late_delivery( _arg1, _arg2):
import pandas as pd
row = {'shipping scheduled': _arg1,
'country_str':_arg2}
#Convert it into a dataframe
test_data = pd.DataFrame(data = row,index=[0])
from sklearn import preprocessing
le = preprocessing.LabelEncoder()
test_data['country_str'] = le.fit_transform(test_data['country_str'])
#Predict the late delivery probabilites
predprob_late = random_forest_l.predict_proba(test_data)
#Return only the probability
return [probability[1] for probability in predprob_late]
#Deploying
client.deploy('fraud_predictor5', fraud_predictor5,'fraud_predictor probability',override = True)
client.deploy('late_delivery', late_delivery,'late_delivery_prop',override = True)
