Shiny 2.6

Loading...

En provenance du cours de Johns Hopkins University

Developing Data Products

1703 notes

Johns Hopkins University

Developing Data Products

1703 notes

Cours 9 sur 10 dans Specialization Data Science

A data product is the production output from a statistical analysis. Data products automate complex analysis tasks or use technology to expand the utility of a data informed model, algorithm or inference. This course covers the basics of creating data products using Shiny, R packages, and interactive graphics. The course will focus on the statistical fundamentals of creating a data product that can be used to tell a story about data to a mass audience.

À partir de la leçon

Shiny, GoogleVis, and Plotly

Now we can turn to the first substantive lessons. In this module, you'll learn how to develop basic applications and interactive graphics in shiny, compose interactive HTML graphics with GoogleVis, and prepare data visualizations with Plotly.

Rencontrer les enseignants

Brian Caffo, PhD
Professor, Biostatistics
Bloomberg School of Public Health
Jeff Leek, PhD
Associate Professor, Biostatistics
Bloomberg School of Public Health
Roger D. Peng, PhD
Associate Professor, Biostatistics
Bloomberg School of Public Health

In this final example, we're going to show how to use the brush argument in the UI

and the brushedPoints argument in the server to generate an interactive graphic.

And in this case what we're going to do is highlight points and then

show in a scattered plot, and then show the fitted line just for those points.

Okay, so this is a nice example of interactive graphics which I think is one

of the key functionalities of Shiny and one of its primary uses for

those of us that are more on the analysis side.

Okay, so let's get right.

Okay, so once again I've copied the code from the R mark down document and

I'm just going to go through the various lines, and

then hopefully you'll get the gist, but I'd like you at home also to try it out.

It only takes a second to copy the code and run the text, and

I think you'll see a lot more by doing it yourself, than by having me describe it.

So once again we're going to have the standard layout, side bar and

in the side panel it's going to have a display of the Slope,

slopeOut, Intercept, intOut.

So the textOutput (intOut), we're going to be looking for

that label from the server function, textOutput (slopeOut),

we're going to looking for that label from the Shiny server.

Okay, so we got to remember to have both those, otherwise they won't display right.

And then we want plotOutput, plot1.

We got to make sure that we're getting the plotOutput,

that we're creating a plot in the server labeled plot1.

And here is the id brush = brush1.

That label we're going to need to use that on the server side.

And this brush = and then brushOpts for the brush options.

Okay, now let's look at the server.

So the Shiny server, input, takes in input and output lists, okay?

Now, it's going to create a model.

The model is going to be reactive.

It's going to be reactive to the data, okay?

So this is going to be reactive as we input new things from the user interface,

this is going to change, okay?

So, and remember it has this notation of being inside the curly braces

also inside the parentheses for the reactive function, okay?

brushed_data is just going to be short hand for

grabbing the data from brush1, okay?

So here's our input, input$brush1, okay?

And our brush points are going to be from the dataset, trees, okay?

The xvar is going to be Girth, yvar is going to be Volume, okay?

And this will create a little data frame with Girth and Volume.

And if it's null, I'm sorry, if there's fewer than two points,

then it's just going to return null, okay?

And then otherwise, It's going to fit a linear,

it fit in return a linear model where Volume is the y variable,

Girth is the x variable, and the data here is that brushed_data, okay?

And then we don't need to put a parentheses here because this is inside

the reactive statement, okay?

The data is going to be that brushed_data and this will output that linear model.

Now, if that model is fit or output$slopeOut,

remember we're looking for some text with that label.

Okay, here's our renderText telling Shiny that this is going to be text,

this is going to be rendered in the user interface.

We've got our curly braces, okay?

And if it's null then it's just going to say, No Model Found.

Otherwise it's going to say, it's going to grab the slope term.

Okay, and then same thing for the intercept.

So the same discussion but it's going to grab the intercept term, okay?

And here's plot1 which we need to display.

And that's going to be renderPlot, okay?

And it has the reactive notation of having the curly braces there inside the function

argument.

And it's going to be plot(trees$Girth, trees$Volume).

So, Girth is x, Volume is y, it's got nice x labels and y labels, okay?

It's got a title, it says Tree Measurements.

Let's go down here and just see if the model is not null,

then it's going to add a line.

Just now, because it's through the model, right?

Through the model, that was fit just on the point scribe from the brush.

It's going to fit the line, it's going to plot the line that is only for

those specific data points.

Okay, great. All right, now,

let's run it and see how it looks.

Okay, I ran the app.

Okay, so now see no model has been fit, so it is correctly saying,

No Model Found for the Slope, and No Model Found for the Intercept.

Okay, so if I highlight this cluster of points,

that’s the line that you would fit and it gives you the slope and the intercept.

But if I were to grab all of those, that's the line that and that's the intercept.

Let's grab over here where there's no point, see what happens.

Okay, it correctly says, No Model Found and No Model Found.

And if you do the whole thing, everything except that last point, there it goes.

Let's see if we can get the whole thing, there we go.

That would be the line that you'd fit if you grabbed all the data points.

Okay, so this is a nice little interactive graph that you can create using Shiny.

And I'd like you to go ahead and try to recreate this example, but

then also maybe try to create your own example of interactive graph using one of

R's default, kind of canned data sets.

And I think this is one of the primary uses for Shiny.

And I think at this point,

you should also have enough to go on to start making Shiny useful for you.

And there's a lot of other things you can do with Shiny, such as input and

output data sets, data frames.

You can actually embed RGL, RS 3D graphics library in a Shiny webpage using WebGL.

But that stuff's maybe a little bit more fringy, especially the RGL stuff.

Right now, I think you have enough to get started, and I think you should

definitely should give Shiny a try, and work it into your standard toolbox.

Explorer notre catalogue

Rejoignez-nous gratuitement et obtenez des recommendations, des mises à jour et des offres personnalisées.

Coursera propose un accès universel à la meilleure formation au monde, en partenariat avec des universités et des organisations du plus haut niveau, pour proposer des cours en ligne.

© 2019 Coursera Inc. Tous droits réservés.

Télécharger dans l'App Store

Disponible sur Google Play