Preview
Video Player
00:00
00:00
Our Data Set - Flower Power
4:07An introduction to the dataset we'll be using in this stage of the course, the Iris Data Set.
[MUSIC]
0:00
So far,
we've generated graphs with lists of data.
0:04
That works pretty well.
0:08
But we really can't expect,
0:10
all of the data we want to explore to be
in the form of Standard Python lists.
0:11
Matplotlib relies heavily on
NumPy arrays as a data type.
0:16
We won't cover NumPy arrays
in detail in this course, but
0:21
I've included links to further
resources in the teacher's notes.
0:23
For this course, we'll be converting
our data into Python lists, and
0:27
we'll see how NumPy can be
leveraged in a future course.
0:31
Now, the Tree House Studio, where I'm
at right now, is in Portland, Oregon.
0:35
Approximately, one hour south of Portland,
0:39
you'll find yourself in the heart
of Oregon's Iris farms.
0:41
In fact, every May,
0:44
there's an entire festival dedicated
to the humble iris flower.
0:46
Other than a bit of trivia, what does
that have to do with data visualization?
0:50
Well, in 1936, a British statistician and
biologist named,
0:55
Ronald Fisher, used a dataset from Edgar
Anderson on variations in iris flowers.
0:59
This has become a popular dataset
to explore and can make for
1:05
some interesting visualizations.
1:08
The dataset takes a look at three species
of iris flowers, and the length and
1:10
width of their flowers,
sepals, and petals.
1:14
For those who have forgotten, or
never took botany, the sepals
1:18
typically function as protection for
the flower when it's in bud stage, and
1:21
offer support for
the petals when the flower is in bloom.
1:25
The petals surround the reproductive
portion of the flower and
1:28
are often designed to attract pollinators.
1:30
In the case of irises, the sepals and
petals are very distinctive.
1:33
The dataset itself is relatively small,
with only 150 samples.
1:38
This will be great for our purposes,
as it will allow us to generate
1:42
a variety of charts without providing
an overabundance of data points.
1:46
But it's large enough for us to work
with several different charts, and
1:50
see how they work from
a reporting standpoint.
1:54
We're going to take a look at this dataset
with three different charts to see what,
1:57
if any conclusions we can
make with our Iris dataset.
2:00
Let's briefly look at our dataset and
prepare it for matplotlib.
2:05
Let's start by looking at
the iris_source.txt file.
2:09
It provides the information
about the data source and
2:14
shows us the attributes
of the data in iris.csv.
2:18
Since that file doesn’t have a header row,
this text file’s a great resource.
2:21
The columns are sepal length,
sepal width, petal length,
2:25
petal width, and the botanical
class to which the iris belongs.
2:30
If we open up iris.csv,
2:35
we see the 50 different
2:40
samples of each class.
2:44
So let's go back and
create a new notebook.
2:53
We'll need our imports.
3:00
We'll be using the csv
module to read our file.
3:02
So we mean to import that.
3:05
Now, we need our input filename and path.
3:10
Let's assign it to a variable
to make it easier to change and
3:12
provide more readable code.
3:15
Finally, we're reading
the rows from our csv file and
3:24
append them to a list, then we'll
be ready to tackle some data vis.
3:27
With open(input_file, 'r')
3:34
We'll import that, as iris_data.
3:41
We'll make our list irises
from our csv.reader.
3:45
Now, that we have our data in a usable
data structure for matplotlib,
3:56
let's take a short break.
4:00
In our next video,
we'll start doing some data visualization,
4:01
and we'll see what conclusions
bloom from our data.
4:04