Pandas

Question 1

Dataframes are the pandas equivalent of a Numpy 2D ndarray, with a few key differences

Answer 1

Axis values can have string labels, not just numeric ones.

Dataframes can contain columns with multiple data types: including integer, float, and string.

Question 2

Read CSV with pandas

Answer 2

import pandas as pd

f500 = pd.read_csv(‘f500.csv’,index_col=0)

Question 3

DataFrame.shape attribute to assign the shape of f500 to f500_shape.

Answer 3

f500_shape = f500.shape

Question 4

Use Python’s type() function to assign the type of f500 to f500_type.

Answer 4

f500_type = type(f500)

Question 5

se the head() method to select the first 6 rows of f500. Assign the result to f500_head.

Answer 5

f500_head = f500.head(6)

Question 6

Use the tail() method to select the last 8 rows of f500. Assign the result to f500_tail.

Answer 6

f500_tail = f500.tail(8)

Question 7

Use the DataFrame.info() method to display information about the f500 dataframe.

Answer 7

f500.info()

Question 8

Select the industry column of f500. Assign the result to the variable name industries.

Answer 8

industries=f500[“industry”]

Question 9

In order, select the revenues and years_on_global_500_list columns. Assign the result to the variable name revenues_years.

Answer 9

revenues_years=f500[[“revenues”, “years_on_global_500_list”]]

Question 10

In order, select all columns from ceo up to and including sector. Assign the result to the variable name ceo_to_sector.

Answer 10

ceo_to_sector=f500.loc[:,”ceo”:”sector”]

Question 11

By selecting data from f500:
Create a new variable toyota, with:
Just the row with index Toyota Motor.
All columns.

Answer 11

toyota=f500.loc[“Toyota Motor”]

Question 12

By selecting data from f500: Create a new variable, drink_companies, with:
Rows with indicies Anheuser-Busch InBev, Coca-Cola, and Heineken Holding, in that order.
All columns.

Answer 12

drink_companies=f500.loc[[“Anheuser-Busch InBev”, “Coca-Cola”, “Heineken Holding”]]

Question 13

By selecting data from f500: Create a new variable, middle_companies with All rows with indicies from Tata Motorsto Nationwide, inclusive.
All columns from rank to country, inclusive.

Answer 13

middle_companies=f500.loc[“Tata Motors” : “Nationwide”, “rank”: “country”]

Question 14

We’ve already saved a selection of data from f500 to a dataframe named f500_sel.

Find the counts of each unique value in the country column in the f500_sel dataframe.
Select the country column in the f500_sel dataframe. Assign it to a variable named countries.
Use the Series.value_counts() method to return the value counts for countries. Assign the results to country_counts.

Answer 14

countries=f500_sel[“country”]

country_counts=countries.value_counts()

print(country_counts)

Question 15

From the pandas series countries_counts:

Select the item at index label India. Assign the result to the variable name india.

Answer 15

countries = f500['country']
countries_counts = countries.value_counts()

india = countries_counts.loc[“India”]

Question 16

From the pandas series countries_counts: In order, select the items with index labels USA, Canada, and Mexico. Assign the result to the variable name north_america.

Answer 16

north_america= countries_counts[[“USA”,”Canada”,”Mexico”]]

Question 17

By selecting data from f500:

Create a new variable big_movers, with:
Rows with indices Aviva, HP, JD.com, and BHP Billiton, in that order.
The rank and previous_rank columns, in that order.

Answer 17

big_movers = f500.loc[[“Aviva”, “HP”, “JD.com”, “BHP Billiton”], [“rank”,”previous_rank”]]

Question 18

By selecting data from f500:

Create a new variable, bottom_companies with:
All rows with indices from National Gridto AutoNation, inclusive.
The rank, sector, and country columns.

Answer 18

bottom_companies = f500.loc[“National Grid”:”AutoNation”, [“rank”,”sector”,”country”]]

Question 19

in f500 Subtract the values in the rank column from the values in the previous_rank column. Assign the result to rank_change.

Answer 19

rank_change=f500[“previous_rank”]-f500[“rank”]

Question 20

Use the Series.max() method to find the maximum value for the rank_change series. Assign the result to the variable rank_change_max.

Answer 20

rank_change =  f500["previous_rank"] - f500["rank"]
rank_change_max = rank_change.max()

Question 21

Use the Series.min() method to find the minimum value for the rank_change series. Assign the result to the variable rank_change_min.

Answer 21

rank_change_min = rank_change.min()

Question 22

Return a series of descriptive statistics for the rank column in f500.
Select the rank column. Assign it to a variable named rank.
Use the Series.describe() method to return a series of statistics for rank. Assign the result to rank_desc.

Answer 22

rank = f500["rank"]
rank_desc = rank.describe()

Question 23

Use Series.value_counts() and Series.loc to return the number of companies with a value of 0 in the previous_rank column in the f500 dataframe. Assign the results to zero_previous_rank.

Answer 23

zero_previous_rank = f500[“previous_rank”].value_counts().loc[0]

Question 24

Use the DataFrame.max() method to find the maximum value for only the numeric columns from f500 (you may need to check the documentation). Assign the result to the variable max_f500.

Answer 24

max_f500 = f500.max(numeric_only=True)

Question 25

f500_desc = f500.describe()

Answer 25

Return a dataframe of descriptive statistics for all of the numeric columns in f500. Assign the result to f500_desc.

Question 26

The company “Dow Chemical” has named a new CEO. Update the value where the row label is Dow Chemical and for the ceo column to Jim Fitterling in the f500 dataframe.

Answer 26

f500.loc[“Dow Chemical”,”ceo”] = “Jim Fitterling”

Question 27

Create a boolean series, motor_bool, that compares whether the values in the industry column from the f500 dataframe are equal to “Motor Vehicles and Parts”

Answer 27

motor_bool = f500[“industry”] == “Motor Vehicles and Parts”

Question 28

Use the motor_bool boolean series to index the country column. Assign the result to motor_countries.

Answer 28

motor_countries = f500.loc[motor_bool, “country”]

Question 29

Use boolean indexing to update values in the previous_rank column of the f500 dataframe:
There should now be a value of np.nan where there previously was a value of 0.

import numpy as np
prev_rank_before = f500[“previous_rank”].value_counts(dropna=False).head()

Answer 29

f500.loc[f500[“previous_rank”] == 0, “previous_rank”] = np.nan

Question 30

Create a new pandas series, prev_rank_after, using the same syntax that was used to create the prev_rank_before series.

Answer 30

prev_rank_after = f500[“previous_rank”].value_counts(dropna=False).head()

Question 31

Add a new column named rank_change to the f500 dataframe by subtracting the values in the rank column from the values in the previous_rank column.

Answer 31

f500["rank_change"] = f500["previous_rank"] - f500["rank"]
rank_change_desc = f500["rank_change"].describe()

Question 32

industry_usa = f500[“industry”][f500[“country”] == “USA”].value_counts().head(2)

Answer 32

sector_china = f500[“sector”][f500[“country”] == “China”].value_counts().head(3)