Intro to Python, Class 4 Solutions

Challenge-scatterplot

(p9.ggplot(data=smoke_complete,
         mapping=p9.aes(x="age_at_diagnosis", y="years_smoked"))
    + p9.geom_point(mapping=p9.aes(color="gender"))
    )

Challenge-comments

# create object holding data and aesthetic mapping
my_plot = p9.ggplot(smoke_complete, p9.aes(x='vital_status', y='cigarettes_per_day'))
# plot using object
(my_plot 
    # boxplot
    + p9.geom_boxplot(outlier_alpha=0)
    # points overlaid on boxplot
    + p9.geom_jitter(alpha=0.2, color='purple')
)

This allows the coder to experiment with different ways of representing the data, while holding the data and comparisons (e.g., axes) static.

Challenge-layers

my_plot = p9.ggplot(smoke_complete, p9.aes(x='vital_status', y='cigarettes_per_day'))
(my_plot 
    + p9.geom_jitter(alpha=0.2, color='purple')
    + p9.geom_boxplot(outlier_alpha=0)
)

Yes, the order matters: later layers are plotted on top of previous layers.

Challenge-violin

(p9.ggplot(data=smoke_complete,
          mapping=p9.aes(x="vital_status", y="cigarettes_per_day"))
     + p9.geom_jitter(alpha=0.2, color="orange")
     + p9.geom_violin()
    )

Challenge-line

# create new data object grouping by year of birth and gender
gender_counts = birth_reduced.groupby(["year_of_birth", "gender"])["gender"].count()
# reset index to format data frame
gender_counts = gender_counts.reset_index(name="counts")
gender_counts.head()
# plot graph
(p9.ggplot(data=gender_counts,
          mapping=p9.aes(x="year_of_birth", y="counts"))
     + p9.geom_line(mapping=p9.aes(color="gender"))
    )

Challenge-dashes

(p9.ggplot(yearly_counts)
    + p9.geom_line(p9.aes(x='year_of_birth',
                          y='counts',
                          linetype='disease'))
    )

Challenge-panels

(p9.ggplot(data=gender_counts,
          mapping=p9.aes(x="year_of_birth", y="counts"))
     + p9.geom_line(mapping=p9.aes(color="gender"))
     + p9.facet_wrap("gender")
    )

Challenge-axis

Search “tick marks ggplot2”, most results will be for R and may need to be adapted! One possible solution: https://rstudio-pubs-static.s3.amazonaws.com/3364_d1a578f521174152b46b19d0c83cbe7e.html

(p9.ggplot(data=gender_counts,
          mapping=p9.aes(x="year_of_birth", y="counts"))
     + p9.geom_line(mapping=p9.aes(color="gender"))
     + p9.theme(axis_ticks_major_y=p9.element_blank()) #removes y axis ticks
     + p9.facet_wrap("gender")
    )