Wrap up

Putting it together

This lesson shows what the beginning of an analysis might look like. Generally, data analysis is done in notebooks, like this one. In a notebook, you can alternate between blocks of code and narrative text.

The first part of an analysis is often importing tools you will need for the analysis. For example, verbs like filter and mutate are imported from siuba.

The imports for this analysis are shown below.

# this is a special object, we use for blanks in exercises!
from setup import ___

# here we import verbs like filter, arrange, and mutate from siuba.
# the import * means to import all of siuba's verbs.
from siuba import *

# here we import everything for plotting from plotnine (like ggplot())
from plotnine import *

# here we import the data for the course
# note that rather than using * to get everything, you can name
# specific things to import (like track_features)
from music_top200 import music_top200, track_features

Exercise 1:

For the artist with the top track in Spain, what country has the most streams for one of their tracks?

Note: you may need to write and run code multiple times.

hint

First, find the artist in the top position in Spain. After, can you get only that artists tracks? Once you do that you should be close!

# getting most streamed track for top artist in Spain
(


)
()
# Note: I would run the pipe with...
#   * only the commented out filter first, to get the artist (KAROL G)
#   * then, with the uncommented filter and arrange
(music_top200 
  # >> filter(_.country == "Spain")
  >> filter(_.artist == "KAROL G")
  >> arrange(-_.streams)
)
country position track_name artist streams duration continent
8204 Mexico 5 Tusa KAROL G 4830583 200.96 Americas
3600 Spain 1 Tusa KAROL G 3295893 200.96 Europe
7897 United States 98 Tusa KAROL G 2385525 200.96 Americas
... ... ... ... ... ... ... ...
3563 Estonia 164 Tusa KAROL G 4681 200.96 Europe
7265 Malta 66 Tusa KAROL G 3732 200.96 Europe
8095 Cyprus 96 Tusa KAROL G 2168 200.96 Asia

40 rows × 7 columns

Exercise 2:

Subset to keep only tracks in Hong Kong, then calculate a new column called stream_seconds, that’s equal to streams times their duration.

Once you’ve done that, try deleting the comments (#) in the code below to plot the data.

(music_top200
  >> ___
  >> ___
  #>> ggplot(aes("stream_seconds", "position", color = "duration"))
  # + geom_point()
)
⚠️: Don't forget to replace all the blanks! 
(music_top200
  >> filter(_.country == "Hong Kong")
  >> mutate(stream_seconds = _.streams * _.duration)
  >> ggplot(aes("stream_seconds", "position", color = "duration"))
   + geom_point()
)