Pandas

Question 1

Axis?

Answer 1

Axis 0 = rows

Axis 1 = columns

Question 2

Reorder columns in dataframe
Option 1)
Option 2)

Answer 2

1) df.sort_index(axis=1)

2) df.reindex(columns=sorted(df.columms))

Question 3

Explore document
1) first rows
2)
3)

Answer 3

Df.head()
Df.info()
Df.describe()

Question 4

Reorder dataframe

1) by index
2) by a particular column

Answer 4

1) df.sort_index()

2) df.sort_values(by=column)

Question 5

Remove columns from dataframe

Answer 5

Df.drop([col1, col2], axis=1)

Question 6

Filter dataframe by a value in a column?

Answer 6

Df[df[‘column’]<5]

Question 7

Filter dataframe by several conditions

Answer 7

Df[df[‘column’] >0 and df[‘column] == ‘Berlin’]

Question 8

Filter column in df by list of values

Answer 8

List = [‘one’, ‘two’]

Df[df[‘col’].isin(list)]

Question 9

Filter column values by contains

Answer 9

Df[df[‘col’].str.contains(pattern)]

Question 10

Unique values in columm

Answer 10

Df[‘col’].unique()

Question 11

Filter column values by does not contain?

Answer 11

Df[df[‘col’].str.contains(‘blabla’)]

Question 12

Join dataframes

1) merge
2) join

Answer 12

1)
Pd.merge(left, right, how=’inner’,on=None,left_on=None, right_on=None
…)

Question 13

Concatenate datasets

Answer 13

Pd.concat([df1,df2],axis=0)

Question 14

Create new columns based on existing columns?

1) 2 conditions
2) 2 conditions

Answer 14

1) ifelse
Np.where(condition, ‘yes’, ‘no’)

2) ternary expression
Df[‘col’] = df[‘number’].apply(lambda x:
‘more than 5’ if x > 5 else ‘5 or less’)

Question 15

Create column based on existing column

1) one column, 3+ conditioma

Answer 15

Df[‘ncol’] = df[‘oldcol’].apply(function)

Question 16

Create new column based on multiple columns

Answer 16

Df[‘newcol’] = df.apply(lambda row: function(row), axis=1)

Question 17

Convert x to string
1)
2)

Answer 17

.astype(str)

Str()

Question 18

Check missing values in column

Answer 18

Isnull().any()

Question 19

Remove rows with missing values

Answer 19

.dropna()

Question 20

Fill missing values

Answer 20

.fillna(0)

Question 21

Filter column values by regex

Answer 21

Df[‘col’].str.contains(pattern,regex=True)

Question 22

Filter groups where the max. value of cumsum_cpo > 0

Answer 22

test_converting = test.groupby(‘campaign_id’).filter(lambda g: g[‘cumsum_cpo’].max() != 0)

Question 23

Cumsum by groups

Answer 23

facebook_costs[‘cumsum’] = facebook_costs.groupby(‘campaign_id’)[‘spend’].cumsum()

Question 24

select index by condittion and convert to list

Answer 24

df.index[conversions_sitelinks[‘conversions’].str.len() > 7].tolist()

Question 25

Find maximum values per group

Answer 25

result = df.groupby(‘A’).apply(lambda g: g[g[‘B’] == g[‘B’].max()])

Question 26

Count of values

Answer 26

Value_counts()

Question 27

Count of values, relative %

Answer 27

Value_counts(normalize=True)[value]

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

Question 28

group by and aggregate

Answer 28

data[data[‘item’] == ‘call’].groupby(‘month’).agg(
# Get max of the duration column for each group
max_duration=(‘duration’, max),
# Get min of the duration column for each group
min_duration=(‘duration’, min),
# Get sum of the duration column for each group
total_duration=(‘duration’, sum),
# Apply a lambda to date column
num_days=(“date”, lambda x: (max(x) - min(x)).days)

Question 29

Correlation two columns

Answer 29

df[‘col1’].corr(df[‘col2’])

Question 30

Percent change to previous row

Answer 30

• datetime as index

pct_change()

Question 31

Iterate through two lists at once

Answer 31

>>> letters = ['a', 'b', 'c']
>>> numbers = [0, 1, 2]
>>> for l, n in zip(letters, numbers):
...     print(f'Letter: {l}')
...     print(f'Number: {n}')

Question 32

Create new column with percentage of total from another column

Answer 32

df[‘col_pct’] = df.col / df.col.sum()

Question 33

Print dimensions of df

Answer 33

print(f’df:{df.shape}’)

Question 34

Number of unique values in a column

Answer 34

len(set(df.col))

Question 35

Correlation with all other columns in df

Answer 35

result = df.corr()[[‘col’]]

Question 36

Join dataframes (merge)

1) index
2) other columns

Answer 36

result = df1.merge(df2, on='index', how='left')
result = df1.merge(df2, left_on='col1', right_on='col2', how='left')

Question 37

Display information to a function

Answer 37

?name

Question 38

display version of pandas

Answer 38

pd. __version__

Question 39

Set max. number of rows

Answer 39

pd. set_option(‘display.max_rows’, 500)

pd. options.display.

Question 40

Remove duplicates from list:

1) list comprehension
2) set

Answer 40

1) res = []
[res.append(x) for x in test_list if x not in res]
2) list(set(test_list))

Question 41

Unique values in a column

Answer 41

set(df.col)

Question 42

rename column in dataframe

Answer 42

df.rename(columns={‘two’:’new_name’}, inplace=True)

Question 43

replace string in a column with another string

Answer 43

df[‘newcol’] = df[‘oldcol’].str.replace(‘str’,’another_str’)

Question 44

stack two dataframes

Answer 44

transactions = pd.concat([transactions_1, transactions_2],axis=0)

Question 45

new column: sum over groups

Answer 45

final_new[‘sum’] = final_new.groupby(‘created’)[‘count_transactions’].transform(“sum”)