We’re going to start by clicking out Aviation Statistics BISM data source. That will launch into a blank Power View report.
This report will help me decide how I can get to New York. This is an exercise that I do frequently as I travel; you might as well. I’m going to begin by applying filters to the entire view.
I know which airport I’ll be leaving from. For me that’s Detroit, which is DTW.
I know where I’m going. I don’t know exactly which airport I want to go to, but I know I want to go to New York state. To narrow that down to the city a little better I’ll choose a hub size of “Large”, because I know the large hubs are around New York City.
These are filters that will apply to anything I put into the view.
The first thing I’m going to decide is when I’m going to go to New York. I want to minimize the possibility of delays. I’d like the least amount of delay in my trip. I have an attribute called “MonthString” that I’ll select.
That has the month and year. If I scroll down to my historical flights, I can find a measure called On- Time Flight Percent, which I’ll be using a lot. This gives me the percentage for the last two years. But that’s not really what I want. I want a line chart. So I’ll just click on “Line” to change it into a line chart.
Then I’ll put that chart down at the bottom of the screen. This represents the overall (all airlines, all flights) on-time percentage by month over the last two years. I can see it’s actually been in an uptrends- -which is great.
I have a pretty good chance of getting a flight that’s on-time, but still at 70% it’s not that impressive. I wonder if I fly a certain day of the week if I’ll have better luck. I do have an attribute or that, so I’ll drag in the day of week as a string. And again I’ll pull in on-time flight % and add that to the chart.
This time a column chart would work really well. That shows me by day of the week what’s the best day to fly. I want to sort tis by on-time flight % in descending order. This shows me that Saturday, Sunday and Tuesday are the three best days of the week to travel.
As I click on these bars you can see the line chart is changing because Power View will automatically cross-filter between charts. If you click on one chart the other will be filtered.
Over time Saturdays have been getting a lot better, while Thursdays have stayed pretty bad.
Clearly the weekend is the best time to get an on-time flight. I’m not sure I want to travel on the weekend, though. I’ve established that I’m probably flying on a Tuesday. Let’s see what carrier I’d like to take.
I’ll drag the carrier name into the surface. Again I’ll use my on-tie flight % as the primary driver measure. So I can see the breakdown of on-tie flight % by carrier on Saturdays. Let’s see where we’re going to be going. If I click on the dimension list I can drag in the Airport column.
I’ll put that on the top of the list, and if I sort by the on-time flight % it looks like I’m going to be flying into LaGuardia either on American Eagle or Delta. Let’s narrow that down more and see what flight number I’d like to take. Let’s bring in the average time the flight’s been leaving as well.
If I take a step back and see what I have in this view now, you can see that if I click on my days, I get the on-tie flight % on those days. You can see that Tuesday is a good day to fly, but it’s not so good if I’m flying in September or April or January. But most of the other months are pretty good.
What If I wanted to fly on Tuesday? These would be the best flights to take.
There’s a quick example of using Power View for some personal data exploration. We were able to knock together that analysis pretty quickly and make some decent decisions on the month, day and destination airport–even which flight number and time–that would be the best to select when traveling to New York while minimizing flight delays.