Before we get to our data source,

let's familiarize ourselves with some concepts of the Bokeh Library.

Some of the common imports we'll use and some terminology in Bokeh.

We'll start by generating a basic plot with some Python lists.

Simulated showcase, our tools, and some options.

To get started, we need to have our libraries available.

I've included a requirements.txt file that has the libraries we'll be using for

this course.

Download and unzip that file, and

install the requirements by running the following

python3 -m pip install -r requirements.txt.

Once we have those libraries set up and installed,

we can start a new project file in Python.

Stage1_2.py.

Let's begin with our imports, and talk briefly about them.

From bokeh.io, import output file and show.

And from bokeh.plotting, And import figure.

Great, output file defines the name of our output file

to be generated when the show method.

When the show method is called, it will generate the HTML and

display it in a browser.

There are other possibilities for output, such as to a Jupiter notebook, but

that is beyond the scope of this course.

Jupiter notebooks are amazing though, and you should check the links

in the teacher's notes to learn more about those later.

The next import, figure, allows us to create figure objects for plotting and

has options for things such as plot width, height, tool selection, and many more.

But what is a figure object?

It is where our visualization will be drawn.

It is essentially Bokeh's built in graph paper if you will, and provides us

with some default access grids and tools that will simplify plot creation.

Let's define our output file and name it test.html.

Then, we'll set up a basic figure object and call it plot.

That is 600 pixels wide

600 pixels tall,

And includes a few of Bokeh's built-in tools for figures.

These tools will be used in the browser which we'll see and

talk about in a little bit.

Now, we need something to plot or visualize, right?

We can pass into our plot, a variety of data types.

Things like lists, arrays, and panda's data frames.

Really, anything that is a sequence can be passed in as x and y coordinates.

One thing to remember here is that the sequences must be of the same length.

If you pass in five values for x, you will need to pass five value for y.

We also have several options for the shape of our plot point or glyph.

Let's plot some points with a square shape of size 20.

Let's choose square for x values,

lets do (x=[1, 2, 4, 8, 10]).

Our y values, Can be ([6,

2, 18, 4, 9]).

And we want them to have a size, 20.

These are a sequence of x and y pairs.

So the first point will be 1, 6, the second, 2, 2 and so on.

Now, we just need to pass our plot into the show method and call it, like so.

Then we can run our script.

And there it is, automatically in the browser.

Now, depending on your version of your operating system, you may see an execution

error about not being able to understand the open location message.

Don't worry about this if you see it.

So long as the plot loads correctly, your code works.

You will notice, off on the upper right-hand side,

the tools we requested when configuring our plot, Pan, Box Zoom, and Reset.

Pan allows us to move the plot around.

Box Zoom allows us to zoom in on a specific location in the plot.

And Reset, well, it resets the plot.

Try them out on your own to get a feel for them.

I put a link in the teacher's notes for other tools you can include as well.

Now, if we go back and take a look at what was generated by Bokeh,

in our test.html output file, and let Python do a bit of code reformatting

with Control+Alt+L on Windows or Alt+Command+L on Mac.

We see that it generates a standard HTML file with calls to Bokeh's CSS style sheet

and JavaScript on lines seven and nine, respectively.

If we scroll down to around line 36, we see

that Bokeh generated a JSON document which is used to describe our visualization.

We won't spend a lot of time going over this, but I wanted to point out that this

JSON object includes all of the configuration for our plot.

For example, if we look at around line 222,

We'll see our column names and the associated data that we provided.

And if we look at around line 145, We see that our data is of type ColumnDataSource.

As you can see, having Bokeh generate all of this for

us to built our plot is a huge time saver and all done on our end in native Python.

Let's take a short break before we start working with more complex data structures

then basic lists and introduce a powerful Bokeh data object, ColumnDataSource.