Original subtitles

Welcome to this lesson.

So we're going to continue to look at our graphs.

Now, in the previous lessons, we looked at the column in the bar graph and we said that they were

really good.

When you wanted to understand which value was highest, which value was lowest, and how much bigger

one value is than the other.

Now, in this lesson, we're going to be looking at trend analysis.

So what we want to understand is how does your data change over time?

So we're going to be looking at our dates and we're going to be seeing how the values change over that

time.

So we're going to be using our line graph to begin with.

So you can see that the line chart is over here on the visualization.

So let's select that and let's just bring it across, make it nice and big like we have the others.

And you'll see that we've got our x axis and Y axis.

We've also got a secondary y axis that we'll have a quick look a little bit later.

But you can see that the welds are pretty much the same as what we did with the column and bar graph.

So we're going to be using our date, so we're going to be using our data in the x axis.

Just remember that a line chart should always be showing the data and how it changes over time and on

y axis is going to show our value, which in this case we're going to stick with our sales.

Now you will see that when this graph is displayed.

It's actually got quite a lot of data on there at the moment.

And the reason for this is because our audit date has all of that detail.

So currently we've got a year called month and day.

So basically we are showing every day of data from January 2012 through to after July 2014.

So as you can imagine, there's quite a lot of data that we're showing.

But you may remember that we are able to change the way that we drill up and drill down when we're using

a hierarchy.

So if we want to see this at a different level of detail, we can go up and we can use our icons that

we came across earlier in the course.

And you'll remember that we had a drill up button.

So when I drill up, you'll see now that it drills up one level in the hierarchy.

So in this case, now we're looking at our year, quarter and month.

Again, if I go one level more up, we'll just be looking at our year end quarter and I can go one final

level to just show my year.

So by using these icons, I'm able to actually drill up and drill down in terms of the level of detail

that I'm looking at.

You may remember that these two arrows allowed me to jump down the level of the high.

So in this case, if I select the two arrows, I actually show my cells by quarter.

So what this is doing now is actually just showing a graph with quarter and with sales.

And this can be really useful if you're doing seasonality and you've got several years worth of data

and you want to see how does it change over different quarters.

If you took a one level mode down, you would actually see your sales by month.

So again, this is just showing you a different month and what your sales are.

So as I say, that's really good when you're looking at seasonality.

Again, though, we can go back up and we go back to the year.

One that makes more sense, though, is that often you want to see your sales by, say, year end quarter.

So you remember that this expands.

So when we select this now we're looking at some of sales by year end quarter.

And if we go one more level down, we're looking at some of sales by year, quarter and month.

So this is basically our year and month view, which was one that you would probably would use quite

a bit when you're doing your trend analysis.

Okay, so hopefully then all makes sense.

Just remember you're working with the hierarchy with our date and we're then using or drill down and

drill up buttons to be able to look at the level of detail.

Now we're going to keep this on the monthly view.

As you can see, the last month, we've got very low sales versus previous month.

So just just something to look at to take into consideration.

Okay.

Let's have a look at a few other features that we have.

So we didn't mention that we got the secondary y axis.

So what this allows you to do is actually drag another numeric value, say profit, for example, and

drop it in there.

And what you'll see is that you now get a y axis, and the y axis can have a different range on it versus

what is on your existing y axis.

So basically allows you to compare the values of two different items.

Let's take profit out list, use something that is quite a lot smaller.

Let's use or the quantity for example.

Put that into our y axis and now you'll see that these values are quite different to these values.

But now you would be able to compare and to see how your trends compare against each other by using

that y secondary y axis.

Otherwise, if we were to put this order quantity into the y axis, you will see that the values are

so low that they don't do not even register on the graph.

So that's really why you would use your y axis is when you want to be compared to numeric variables.

But one of them is actually a lot less in terms of quantity than what the other values are over here.

Right.

Let's have a look as well at our legend.

So if we use our legend, you remember that from our columns, we then get each of the items, gets

its own line.

So in this case, we have North America, Europe and Asia that we've seen previously.

And again, you can use highlighting.

If you select these, you'll see how it gets highlighted, get some markers placed on it as well, so

you can actually see how these values change.

You also do get the option of small multiples.

We saw this in the columns.

So if you drag regions into small multiples, you will see here now that we actually get each of the

individual graphs being displayed, each of the different regions, which again can be quite useful

when you want to look at the graphs like separately rather than them being on top of each other.

Let's just go back, take that out, and go back to our normal growth.

Now, one of the things I do want to show you is just a couple of features on the formatting.

So if we go to the formatting, you'll see that the x axis changes a little bit.

It's also got a different time.

So you can have continuous, which is what you would traditionally use when you're using dates in this

way or you can have categorical.

Categorical is basically just creating each date as an actual category.

So you can see that the x axis now changes in the way that it's showing this information.

But in this case, I'm going to go back to keeping it as continuous.

So we'll just change it back again when we set default.

There we go.

And you'll see here you also can change your ranges as to which dates you want to use for your range

in there.

Also, again, if you want to change your font style for your titles and for your actual values itself,

y axis, again, you can change your minimum maximum ranges.

So you can see this is very much the same as what we had for the column graph.

Again, you can change your secondary y axis, your legends as well as small multiples, also your grid

lines and your zoom slider.

So we saw all of that when we were looking at the columns.

The ones that I want to focus on are a little bit further down.

So over here you can change your lines, so you can actually change your line style and how that line

is shown.

So at the moment you've got a solid line style, but you could change this to a dashed if you wanted

to.

You could change it to a dotted, could change the stroke width.

You can make it thicker or thinner depending on what you want to show.

We also have an option.

I'm actually going to change the spec to Solid to show a stepped view.

So the step view is often called the skyline.

And you can see it's kind of looks like a city skyline.

So I prefer the traditional view in terms of a line graph, but you can change that as well.

So you can change the line colors as well if you want this to be a different color.

Change that as well.

If you like markers, you can put your markers on.

Change the type of shape of your markers.

Also the size of those markers, the colors of the markets.

So there's quite a bit of information, again that you can change in your markers.

But the one that I wanted to highlight was our data labels.

Now, traditionally where we have a problem with data labels on line graphs is that they label every

point on here.

And what you tend to have then is a lot of detail on your graphs and it makes it very difficult to read

your graphs.

So one of the really nice options we got here in Power BI is something called label density.

So you can decide how much density you want in your label.

So if you only want a few labels, let's say for example, we want to see, say, 20%.

Now you can see there's only a few labels that are on there, and that's really useful to anybody who's

reading this.

They can get an idea from the labels as to what the values are, but you don't have too many values

that are then making the graph hard to read.

So that's just something to take into account, is that label density is quite useful.

The other side that I wanted to highlight with line graphs is also not analytics.

So I'm going to jump across to analytics and you'll see that we've got a couple of new functions in

here that we can use.

So we have a trend line of forecast and the fund anomalies.

Now I'm not going to look at the fund anomalies because our training data is actually very structured.

There's not much you're going to find, but in your own data, it's going to be worth experimenting

with that and playing with it.

What I do want to have a look, though, is what is the current trend line?

So let's turn that on.

And you can see now that a linear trend line is automatically now drawn onto my graph.

And this is quite useful.

If we were to actually put a slicer with this, you can see that your trend line would then obviously

change as we filtered the data.

So if we put our regions, obviously as we change the data, you can see my trend line is obviously

being recalculated each time.

Okay.

So that's one of the functions we can use, is we can actually add a trend line onto our line graph

and you'll see that you can change the color of this.

Like, for example, I would use maybe a not so dark color with this kind of like the dotted option

for showing the lines.

You can decide, do you want to come down series and highlight values on this if you want to experiment

with that.

The other function that I want to show you is the full costing function.

So if we go down and turn full cost on and what that does now is it looks at your data and it now makes

a sort of forecast as to how it sees those values going forward.

And then you'll see that there's a gray area now that has been covered in.

And what that is, is what is your confidence interval?

So if we look at our full cost, it's saying how many units, well, what units do you want to use going

forward?

So you could change this and say, I want to look at months or I want to look at quarters.

In this case, points would equal months in this case.

So we're going to keep that.

I'm going to go ten months into the future.

Now, this is a function I really like.

Now you can see that my last period quite clearly is not a full month of data.

It's much lower than the others.

So I'm going to go back and I'm going to ignore the last data point.

So we're actually now going to calculate the forecast from here, which makes much more sense because

this is not a full month of data.

Also, you can change how much confidence you want in this.

Obviously, the less confidence you put in there then these gray lines are not going to be as wide.

And basically you click apply on this and you can see now that it's now re calculates back from one

last month and you can then see the forecasted values and you can also see what the confidence interval

is.

Now one of the things that this does do is that if you go back to your shows table is you'll now see

that there's full cost values and confidence.

Values have now been added into your table.

So you could actually now export this as a CSV, take it into your Excel and work with that.

So quite a nice function there.

Right.

So that is our forecast.

And again, you can see that you can change your forecast line colors, your confidence bands, how

you want it to be done, even your tooltip style as well.

So I'm going to leave you to play around with some of those and we're going to conclude this lesson.

I will see you in the next one.