Second

Question 1

Build df from dictionary

Answer 1

pd.DataFrame()

dict = { “country” : [“Nepal”, “India”, “China”], “Capital” : [‘Kathmandu”, “Delhi”, “Bejing”]}

SA = pd.DataFrame(dict)

Question 2

CSV to DataFrame

Answer 2

df = pd.read_csv(‘path/to/dataframe.csv)

Question 3

CSV to DataFrame

1. Index the df

Answer 3

index_col = 0

df = pd.read_csv(‘path/to/dataframe.csv, index_col = 0)

Question 4

Sort from highest to lowest

Answer 4

df.sort_values(‘col’, ascending = False)

Question 5

Sort by multiple variables

Answer 5

df.sort_values([‘col1’, ‘col2’])

1. Sort by the col1
2. Then sort by col2

dogs.sort_values([‘weight_kg’, ‘height_cm’], ascending = [True, False])

Question 6

Subset rows

Answer 6

df[col] > #

0 True
1 False
2 True

Question 7

Subset rows and get data

Answer 7

df[df[‘col’] > #]

0 Bella Labrador Brown
4 Max Labrador Black

Question 8

Subset based on a string

Answer 8

df[df[‘col1’] == ‘string’]

Question 9

Subset based on date

Answer 9

df[df[‘col_date’] > “2015-01-01”]

Question 10

Subset on multiple condiditons

Answer 10

is_x = df[‘col1’] == ‘x’
is_y = df[‘col2’] == ‘y’
df[is_x & is_y]

 col1   col2   number 0    x        y          56

Question 11

Subset using isin()

Answer 11

is_x_or_y = df[‘col1].isin([‘x’, ‘y’])
df[is_x_or_y]

is_black_or_brown = dogs[‘color’].isin([‘black’, ‘brown’])
df[is_black_or_brown]

  name    color       height   0     Bel       brown       88 4     Max      black        55

Question 12

Adding new column + mutating df

Answer 12

df[‘new_col’] = df[‘col’] / 100

Question 13

Individuals per 10K

Answer 13

df[‘per_10k’] = 10000 * df[‘number’] / df[‘total’]

homelessness[‘indv_per_10k’] = 10000 * homelessness[‘individuals’] / homelessness[‘state_pop’]

Question 14

Summary Stats

Answer 14

df[‘col’].mean()

df[‘col’].median()

df[‘col’].mode()

df[‘col’].min()

df[‘col’].max()

df[‘col’].var()

df[‘col’].std()

df[‘col’].sum()

Question 15

Quantile

Answer 15

df[‘col’].quantile()

Where a sample is divided into equal-sized, adjacent subgroups

Question 16

Aggregate summary stats from one column

Answer 16

df[‘col’].agg(‘function’)

df[‘ext price’].agg(‘sum’)

             ext price sum          2.1

Question 17

Aggregate multiple summary stats from multiple columns

Answer 17

df[[‘col1’, ‘col2’]].agg([‘func1’, ‘func2’])

df[[‘price’, ‘quantity’]].agg([‘sum’, ‘mean’])

           price       quantity sum          2.1            25 mean        1                12

Question 18

Cumulative Sum

Answer 18

df[‘col1’].cumsum()

Question 19

Drop duplicates

Answer 19

df.drop_duplicates(subset = ‘col’)

vet_visits.drop_duplicates(subset = ‘breed’)

Question 20

Drop duplicates multiple columns

Answer 20

df.drop_duplicates(subset = [‘col1’, ‘col2’])

vet_visits.drop_duplicates(subset = [‘name’, ‘breed’])

Question 21

Count number of occurrences

Answer 21

df[‘col’].value_counts()

unique_dogs[‘breed’].value_counts()

labrador 4
Chow Chow 3
Poodle 2

Question 22

Count number of occurrences & sort values

Answer 22

df[‘col’].value_counts(sort = True)

Question 23

Count proportions of Total

Answer 23

df[‘col’].value_counts(normalize = True)

Lab 0.5
Chow 0.25
Poodle 0.25

Question 24

Grouped Summary Statistics by attribute Pt2

Answer 24

df[df[‘col’] == ‘attribute’][‘measure col’].stat()

dogs[dogs[‘color’] == ‘black’][‘weight’].mean()

dogs[dogs[‘color’] == ‘black’][‘weight’].sum()
dogs[dogs[‘color’] == ‘brown’][‘weight’].sum()

26
24

Question 25

Grouped Summary Statistics (groupby) Pt2

Answer 25

df. groupby(‘col’)[‘measure col’].stat()
dogs. groupby(‘color’)[‘weight’].mean()

color
black 26
brown 24

Question 26

Multiple stats groupby summary

Answer 26

df. groupby(‘col’)[‘measure col’].agg([min, max, sum])
dogs. groupby(‘color’)[‘weigth’].agg([min, max, sum])

        min       max       sum color black    24         29           53 brown   24        24           48

Question 27

Group by multiple variables

Answer 27

df. groupby([‘col1’, ‘col2’])[‘measure col’].stat()
dogs. groupby([‘color’, ‘breed’])[‘weight’].mean()

color     breed
black     chow           25
              lab               29
              poodle        24
Brown   chow           24
              lab               25

Question 28

Group by multiple cols and multiple measures

Answer 28

df.groupby([‘col1’, ‘col2’])[[‘measure col1’], [‘measure col2’]].stat()

Question 29

Pivot Table

Answer 29

df.pivot_table(values = ‘measure col’, index = ‘group_col’)

By default it takes the mean value of each group

dogs.pivot_table(values = ‘weight’, index = ‘color’)

                 weight color black             26 brown           24

Question 30

Pivot Table: Change the statistic

Answer 30

aggfunc =

df.pivot_table(values = ‘measure col’, index = ‘group_col’, aggfunc = np.median)

Question 31

Pivot Table: Multiple statistics

Answer 31

aggfunc = []

df.pivot_table(values = ‘measure col’, index = ‘group_col’, aggfunc = [np.median, np.mean])

Question 32

Pivot Table: 2 Variables

Answer 32

columns =

df. pivot_table(values = ‘measure col’, index = ‘group_col’, columns = ‘group_col2’)
dogs. pivot_table(values = ‘weight’, index = ‘color’, columns = ‘breed’)

breed chow lab poodle
color
black NA 29 24
brown 24 24 NA

Question 33

Pivot table: Fill Missing Values

Answer 33

fill_value =

df. pivot_table(values = ‘measure col’, index = ‘group_col’, fill_value = 0)
dogs. pivot_table(values = ‘weight’, index = ‘color’, columns = ‘breed’, fill_value = 0)

breed chow lab poodle
color
black 0 29 24
brown 24 24 0

Question 34

Pivot Table: Sum All

Answer 34

df. pivot_table(values = ‘measure col’, index = ‘group_col’, fill_value = 0, margins = True)
dogs. pivot_table(values = ‘weight’, index = ‘color’, columns = ‘breed’, fill_value = 0, margins = True)

breed  chow  lab  poodle  All 
color
black    0        29     24       53
brown  24        24    0         48
All         24        53    24