Gadfly: Simple plotting & Adding Layers

Loading...

From the course by University of Cape Town

Julia Scientific Programming

115 оценки

University of Cape Town

Julia Scientific Programming

115 оценки

This four-module course introduces users to Julia as a first language. Julia is a high-level, high-performance dynamic programming language developed specifically for scientific computing. This language will be particularly useful for applications in physics, chemistry, astronomy, engineering, data science, bioinformatics and many more. As open source software, you will always have it available throughout your working life. It can also be used from the command line, program files or a new type of interface known as a Jupyter notebook (which is freely available as a service from JuliaBox.com). Julia is designed to address the requirements of high-performance numerical and scientific computing while also being effective for general-purpose programming. You will be able to access all the available processors and memory, scrape data from anywhere on the web, and have it always accessible through any device you care to use as long as it has a browser. Join us to discover new computing possibilities. Let's get started on learning Julia. By the end of the course you will be able to: - Programme using the Julia language by practising through assignments - Write your own simple Julia programs from scratch - Understand the advantages and capacities of Julia as a computing language - Work in Jupyter notebooks using the Julia language - Use various Julia packages such as Plots, DataFrames and Stats The course is delivered through video lectures, on-screen demonstrations, quizzes and practical peer-reviewed projects designed to give you an opportunity to work with the packages.

From the lesson

Structuring data and functions in Julia

As a scientific computing language, Julia is well suited to the task of working with data. In this last module, we elaborate on the two most important concepts in Julia, arrays and functions. They are the fundamental building blocks of holding and manipulating data. You should see this week as offering you a chance to further explore concepts introduced in week one and two. You will also be introduced to more efficient ways of managing and visualizing your data. By the end of this module, you will be able to: 1. Apply and understand how to work with arrays 2. Practice Julia functions 3. Explore extension packages 4. be familiar with the Dataframes package 5. Plot a variety of data from the dataset, ready for publication.

Gadfly: Introduction2:21

Gadfly: Simple plotting & Adding Layers7:31

Gadfly: Using Themes8:39

Gadfly: Adding Titles and Axis Labels & Saving a Plot & Importing Data into a DataFrame7:50

Gadfly: Box Plots8:32

Gadfly: Density Plots & Histogram & Violin Plots9:03

Gadfly: QQ Plots & Scatter Plots & Vertical and Horizontal Lines8:30

Gadfly: More Examples5:45

Meet the Instructors

Juan H Klopper
Dr
Department of Surgery
Henri Laurie
Dr
Department of Mathematics and Applied Mathematics

[MUSIC]

The packages that we are going to use in this lecture is, first of all, Gadfly.

That is what this is all about so we are using Gadfly.

You're going to see some error messages,

warning messages pop up from time to time as these packages ungulate self develop.

Now we going to also use Cairo.

Now to be able to export these plots into your local hard drive or into the cloud,

specifically if you want to do PNG files you have to import the Cairo package.

You're also going to use the distribution package because we want to create

random values to use as the data for our plots.

And we're going to take these random values from a variety of distributions.

So we're going to use the distributions package.

And then lastly, Gadfly really is good for

plotting data into work group data, the data frames package is really essential.

Now let's start with some simple plotting.

First of all, let's just create some values to play with.

We going to use the range function,

in these arguments we're going to pass the range values one to five.

If I don't specify size remember it will use the counting values or

integer values so one, two, three, four, and five.

So pick from those elements one, two, three, four, five, five values for me.

And we're going to put that in the computer variable called, xvals.

And there we go we have a five element array which is the column vector

here of 64 bit integer elements, 3, 1, 5, 5, 3.

Every time you run this you're going to get something else.

Let's also create, by the same process, an array of values called y values.

And as I say, we haven't see that the own number generator you're going to get

different values.

Now let's plot these as a scatter plot.

So it's going to take the value three on the x-axis and

then the value three on the y-axis and use a point.

And then it will take x-axis value one, and the y-axis value four, and pluck that.

So here we go, we're just going to say plot and

then we have to refer to these two arrays and that is called the aesthetics.

These values that you pass that will actually be used to create the plot,

those are called aesthetics.

And we take two arguments, x and y.

So x is going to be my array of xvals and

y is going to be my array of yvals and we can plot that to the screen.

Now what you'll notice is some compilation has to take

place invoking the spot function that code has to be compiled so

that it is going to take awhile, depending on the speed of your machine

to have that first block executed if you use the plot function.

Later on you will see the execution

becomes quicker because we've already had compiling of that code.

So there we go remember we had a value of 3 on the x axis and

here we had a 3.0 on the y axis.

And it's going to plot these are, going to do it's best to arrange those values for

you on that plot.

Now these are point markers.

If we go down that is referred to as the geometry lines, points, etc.

And if we don't specify anything and we pass to erase as our aesthetics,

it is going to use the point geometry as it's default but

we can also specify that as an argument value.

And the way to do that is GEOM, short for geometry.point.

So if I were to run that, we're going to get exactly the same thing back.

We're going to have these values plotted as a scatter plot.

You would recognize this as a scatter plot.

Now difficult to see on the screen these marks are quite small.

The color of this, this default color is called deep sky blue and

we'll look at all of this things.

Let's just create something with a few more data point values.

I'm going to create xvals two and yvals two, two computer variables.

They're going to both be fold as arrays of 100 elements each taking values,

integer values in the range from zero through 100 and we're going to plot those.

x and y are static but

this time we're going to call a function on the raise in themselves.

Sort, we'll just sort them, we haven't done anything else so

it's just going to be in ascending order.

So we're going to sort both of those and then create the scatter plot.

And now you're going to see all our little data point values

from zero to 100 both on the x and y axis.

And those are plotted nicely as geom.point geometry.

Now, we can also specify that we want a line drawn as well.

So not only do we want points, a point geometry but a line geometry as well.

So we can do one more than one geometry just in a simple plot.

So again, my site's are going to be x equals those five xvals, and,

five element array and the five element array.

yvals from y aesthetic and I want both point end line.

And that's what we're going to get.

We're going to get the points and they're going to be connected with these lines.

Now I can have another geometry called the smooth geometry.

It actually have two methods.

One is method equals colon align for linear model and

default is if we pass no argument there is just going to be method equals loess.

We can choose linear model here because we want the linear regression here,

a line that indicates line there.

We're going to pass the xvals and the yvals, we want them as a point geometry.

But then we also want geom.smooth which is a type of a line

using the linear models method.

And if we do that, it's going to take some time and its going to draw this line for

us as a representation of the values that we actually have.

At times we want to add a bit more information to our plots.

We want to use two different data sets or

even more data sets to add to the to a single plot.

Now we can do that by just layering these different plots on top of each other.

And for that we use this function as an argument called layer.

So we're still going to call plot and then we're going to have a layer comma,

another layer.

And in inside each of these layers we pass as arguments our z x and our geometry.

So x equals xvals2 y equals yvals2 in the one layer and

in the second layer we again are going to use z x, x is going to be the sort.

Sorting the xvals2 and sorting the yvals2 and there we also want point geometry.

So we can add a third one, comma and have another layer, but let's just do the two,

and we render that and there we go, we see two different scatter plots.

Immediately though, you can see, there's a bit of a problem.

It's quite obvious, they're all in blue.

We don't really know which points belong to which of these layers.

Point being though,

that we can add multiple layers on inside at least the same plot.

Next up we're going to use themes to solve the problem that we have in screen

right now.

[MUSIC]

Ознакомьтесь с нашим каталогом

Присоединяйтесь бесплатно и получайте персонализированные рекомендации, обновления и предложения.

Coursera делает лучшее в мире образование доступным каждому, предлагая онлайн-курсы от ведущих университетов и организаций.

© Coursera Inc., 2018 Все права защищены.

Загрузить из App Store

Загрузить в Google Play