How to create and customise FORECAST LINES in Power BI // Beginners Guide to Power BI in 2021

in this video i'm going to show you how easy it is to add forecast trend lines to your line charts in power bi i'm going to show you how to do it step by step so you can follow along if you want and i'm also going to show you how to tune and tweak your forecast to your liking all of that and more so without further ado let's get started hi my name is fenan and welcome to the solutions abroad youtube channel we're focused on teaching beginners the wonderful world that is power bi i upload new videos every week so make sure you hit that subscribe button and the bell icon to get notified when a new one is out so let me show you how easy it is to do in power bi desktop so let me show you this demo that i prepared for you today this is the calculation groups pbix file that i used to demonstrate how to create calculation groups in power bi if you're interested in how to do that i made a separate video on it but we'll reuse the same data set because it has some information that i want if you don't want to watch that video i want to just give you a quick overview of what the data set contains if we go to the model view you'll see that we have a calendar table which has our date stuff this is for time intelligence we have the orders table which has some information about sales so things like the products how many were ordered and how much for and at what point in time they were purchased and with the time intelligence the calendar and the orders we are able to create some measures that are pretty important so we have the ability to get the cost the profit and sales based on the data that we have but for now we'll focus on the measure sales so let's say for our scenario we want to see how the sales will look like for the next week for the next couple of weeks so what we'll do we'll have to create a new page so we have a new page here we'll put the dates and the sales onto a line chart here [Music] and from here we it has data since back in march last year so maybe we'll just tone that down a little bit so let's say in the last six months instead so that will give us a better sample rate it's not too much and from here to add a forecast line all you need to do is you need to build analytics you'll see under here forecasts you'll just need to click add and you'll see that rbi already starts to add some default numbers for you to start with and now from here you can start tweaking it to make it truly your own kind of forecast so if you focus our view on the analytics pane here for your forecast we have some options here to customize our forecast line and let me explain them to you one by one forecast length controls how far in the future you want to draw your forecast line for so at the moment we have 10 points here which means that because our our visual is a day so each point is a day we are basically saying give me a forecast for the next 10 days right so we can if we want to make that forecast even longer let's say we want for two weeks so we want to do maybe 14 instead for the next two weeks so let me if i hit apply so you'll see that now grows ignore lost points moves your prediction on top of the actuals so it overlays it on top of your actual so you're able to kind of compare and contrast how your predictions are compared to what the actual results are and how far back it goes is based on the points you put here so again because points means days so if we say i want to see how the forecast was for the time for the last seven days if we change this to seven you'll see that the prediction goes back seven points and you'll see that the prediction line on top of the actuals the confidence interval controls how big your upper bound and the lower bound is of your forecast trend so you see um by the gray area on top of your forecast lines so it means how likely the actuals will be within these bounds uh that's the margin of error that your forecast would be the lower the confidence in term intervals the smaller these bounds will be so if we set this to 75 for example you'll see that the bounds also change to become a little bit smaller the seasonality allows you to incorporate trends and sort of seasonal factors to your forecasting line so for example you would use this when you have a full cycle of season in your data so typically for sales you would do a year of sales so you would know when your peak will be maybe you will have a peak in december or maybe you will have a drought of sales in june so a full cycle of that where you have the set of your highs and lows will typically be a year but that can be different depending on what sector or what you're using the data for so in our scenario if we wanted to do one year's worth we need to put 365 points here because each point is a day however as you notice because we don't have a full set of years data we should not be using the seasonality that way because the seasonality works based on sample data right so it will sample based on historical based on historical data in order to use the seasonality effectively it's actually recommended to you have your chart to have four times the data of what you're using as your seasonality so that gives the power bi visual the ability to sample enough seasons to create some more [Music] accurate forecasts so in our case with our sales maybe we can do year but for sure we can do month because we have the past six months so in this case we want to put 31 here or maybe 30. it doesn't really matter so now that will if i hit apply you'll see that the kind of predictions change based on that seasonality so the season which is every month and that's really it for this video i hope it helped you understand how easy it is to start using forecast lines in power bi desktop leave a like on this video if it helped you it's the best way to let me know that you enjoy this type of content get in touch using the social media links that i included in the description box below and thank you so much for watching guys see you again on the next one

Channel: Fernan

Views: 4,101

Keywords: solutions abroad, power bi, powerbi, power bi tutorials, power bi for beginners, beginners guide to power bi, data analytics, dax, data modelling, data visualisation, business intelligence, power bi 2021, power bi guide, power bi forecast, power bi predictions, power bi ai, power bi artificial intelligence, artificial intelligence, power bi forecasting, power bi predicting, power bi confidence interval, power bi seasonality, power bi sampling

Id: MrIwuAVGYR8

Length: 7min 7sec (427 seconds)

Published: Wed Mar 31 2021