Pandas part 2 Review

Python list:

Pandas:

The word "index" refers to the collection of labels for each row.

 

groupby:

Harder Question

• What was the most popular male name during each year in the data?
• What are the three states with the most babies born?

By doing groupby, we can easily approach.

 

# avarage of percent, group by Party 
df['%'].groupby(df['Party']).mean()

# return minimum value, group by Party
df['%'].groupby(df['Party']).min()

# return the size, group by Party
df['%'].groupby(df['Party']).size()

First case

 

# groupby multiple columns
(df['%']
.groupby([df['Party'], df['Result']])
.mean()
)

 

 

Reset Index:

(df
.groupby(['Party', 'Result'])
.mean()
.reset_index()
)
# make index in order

 

Pivot Table:

df.pivot_table(
	index='Party'
    columns='Result',
    values='%',
    aggfunc=np.mena,
)

 

groupby/filter:

isin:

df[df["Party"] == "Democratic") | (df["Party"] == "Republican")] # Ugly
df[df["Party"].isin(["Republican", "Democratic"])] # Better

 

Other:

ca.pivot_table(
	index='Sex', columns='Year',
    values='Name',
    aggfunc=lambda ns: ns.iloc[0])
    .loc['M']
    .value_counts()
    .plot(kind='barh')
)

 

# number of unique woman name!
(ca.pivot_table(
	index='Year', columns='Sex',
	values='Name',
    	aggfunc=len)
.plot()
)