How To Create A Forecast Model In Power BI With Python

In this tutorial, Gaelim is going to show how you can make a forecast model in Power BI using Python. You can utilize the Power BI forecasting feature, which allows you to visually forecast the data you have to as specific day but it will have its limitations. With Python, you can optimize the model a bit more by changing the additive nature of the trend and seasonality, and add predictions into the data set.

***** Video Details *****
00:00 Introduction
00:28 Forecasting samples
00:49 Power BI limitations
01:30 Trend analysis
02:18 Python code
07:22 Power BI implementation
08:10 Python script

***** Learning Power BI? *****

#EnterpriseDNA #Python #PythonTutorial #PythonForPowerBI #PowerBI #PowerBIDesktop #PowerBITutorial #MachineLearning

[gutolima]

totally delivered title's promise in a simple way. good job.

[one_life]

Very much useful with real time project work...Really Appreciate

[dylanarmstrong]

Personally I like to use xgboost when I do regression like this but I've never even heard of the model you used I'll have to do some research.

[Never]

How can we choose the right forecast model that fit our business?

[junior_official_ke]

where can I get the excel file you're using in this video

[Kim-bnub]

Hi, I signed up and have been trying to find the source of dataset but can't find it. can you please help me. thank you

[yufanli]

will this forecast auto refresh after the new data is added in published PBI service? since the python code runs in local env. Do I need to run this locally and update the dashboard?

[beccabruner]

how are your dates formatted in your .xlsx? I keep getting a weird Y axis of years instead of my cost.

[shereenfathima]

What is the source for forecast table where we run the python script

[aosreelakshmi]

AttributeError: 'function' object has no attribute 'forecast'

[Digitalcircus]

I always get hate comments if I say first but I won’t answer my hate comments

[RILZ]

# 'dataset holds the input data for this script
df = dataset
import pandas as pd
df['Date'] = pd.to_datetime(df['Date'])
ts = df.set index ('Date')
ts =ts.asfreq('d')
from statsmodels.tsa.holtwinters import ExponentialSmoothing
train = ts.iloc[:290]
test =ts.iloc [290:]
model = ExponentialSmoothing(train, trend= 'mul', seasonal= 'mul'‚seasonal_periods=7).fit ()
forecast = pd. Dataframe(model, forecast (30))
forecast = forecast.reset_index()
forecast.columns= ['Date', 'pageviews']