Description
1. The dataset “weights.tsv” contains weight values (in pounds) for 20 people: 164, 158,
172, 153, 144, 156, 189, 163, 134, 159, 143, 176, 177, 162, 141, 151, 182, 185, 171,
152.
1.1. Import the data from the file weights.tsv into a Pandas Series object in Python.
(4 points)
1.2. Create a new series object with weights converted to kilograms from pounds (1
pound = 0.453592 kilograms). Round the results to two decimal places.
(4 points)
1.3. Find the mean, median, and standard deviation of both series objects using
Pandas functions. (4 points)
1.4. Plot a histogram of weight (in kilograms) using matplotlib library with 10 bins.
(4 points)
2. For this problem statement, you are given a dataset named “boston.csv”. This dataset
contains information collected by the US Census Service concerning housing in the
areas of Boston, Mass. The data was originally published by Harrison, D., & Rubinfeld,
D.L. (1978). Hedonic prices and the demand for clean air. Journal of Environmental
Economics and Management,5,81–102.
Here is the description of variables/columns in this dataset:
Column name Description
CRIM per capita crime rate by town
ZN proportion of residential land zoned for lots over 25,000 sq.ft.
Page 2 of 4
NDUS proportion of non-retail business acres per town
CHAS Charles River dummy variable
(1 if tract bounds river; 0 otherwise)
NOX nitric oxides concentration (parts per 10 million)
RM mean number of rooms per dwelling
AGE proportion of owner-occupied units built prior to 1940
DIS weighted distances to five Boston employment centres
RAD index of accessibility to radial highways
TAX full-value property-tax rate per $10,000
PTRATIO pupil-teacher ratio by town
LSTAT % lower status of the population
MEDV median value of owner-occupied homes in $1000’s
2.1. Import the dataset “boston.csv” into a Pandas dataframe and obtain the number
of rows and columns for the dataframe. (3 points)
2.2. What is the owner-occupied home value (MEDV) for the lowest nitric oxide
concentration (NOX) from the dataframe? (3 points)
2.3. Create a boxplot of per capita crime rate (CRIM) using Matplotlib. Obtain the
interquartile range for crime rate (CRIM) using Pandas functions. (4 points)
2.4. Subset all columns of the dataframe for rows with outliers of crime rate into a new
dataframe. Compare the mean of AGE between the two dataframes with respect
to crime rate, what do you interpret? (Hint: Outliers exist 1.5 times of interquartile
range above third quartile and below first quartile) (4 points)
2.5. Create scatterplot between distances to employment centers (DIS) and nitric
oxide levels (NOX). Obtain correlation index between the two columns and
interpret their relation. (4 points)
2.6. Similarly, create a scatterplot between highway accessibility index (RAD) and
property tax rate (TAX). Obtain correlation index, compare it to the scatter-plot,
and interpret the relation between RAD and TAX. Take appropriate action on the
data based on your observation. (6 points)
Page 3 of 4
3. We will be using the “tips” dataset from seaborn package for this problem statement.
This dataset contains information about restaurant bills and tips made by people
classified by their gender along with few other attributes which are self-explanatory. You
can import this dataset into a pandas dataframe as follows:
3.1. Calculate percentage of tip amounts for bill totals, rounded to two decimal places
and create a new column “tip_percent” in the same dataframe. (3 points)
3.2. For what days in the week do we have the data, and which day has the highest
bill mean? (Hint: lookup for “groupby” in pandas documentation) (3 points)
3.3. Are there more dinners or lunches? Create a dataframe with this data. Are there
more smokers during lunches or dinners? Create another dataframe with this
data. Join the two dataframes by time of day and calculate the percent of
smokers at lunch and dinner. Compare the results. (6 points)
3.4. Using the boxplot function from seaborn package, create plots on “tip” column for
Male and Female from “sex” column. Compare the boxplots and provide your
interpretation on outliers between males and females. (4 points)
3.5. Create the same boxplots as above for “tip_percent” and “sex”, for tip percent
below 70. Now compare the boxplots between male and female, which boxplot
has more outliers and which one is more symmetric? (4 points)
4. For this last problem statement, you will work on the “avocado.csv” dataset which
contains information related to avocado sales across multiple regions/cities over the
years 2015 to 2018 organised by date. The data contains 10 columns which are self
explanatory.
Page 4 of 4
4.1. Import the dataset file into a Pandas dataframe and identify the count of missing
values per column. Handle missing values based on column type and explain
your reasons behind selecting appropriate techniques. (8 points)
4.2. Convert the fields Type, Year and Region to categorical data type and subset the
dataframe to exclude region “TotalUS” and “West” and sort the dataframe by date
in ascending order. Is the mean price of an avocado higher in 2017 compared to
2016? (4 points)
4.3. Sum up the total volume of avocado sales by region and create a horizontal bar
plot using Matplotlib. Which state from the region has the highest sales of
avocados by volume? Subset the data for that state, create a histogram of mean
price and interpret it. Obtain the correlation index between mean price and total
volume for that state, what do you find? (6 points)
4.4. Provide your observations of the following timeline plot of avocado sales by
volume. Which month consistently has the highest volume of sales every year? In
general, what could be some possible reasons driving this surge in sales? (2
points)
