Weed-Out
Data Visualization - Project
Report
Instructor: Judd Bradbury
Group 11
Aravind Vasu Murugan
Meera Joseph
Ritesh Kunisseri Puliyakote
Sriram Murali
Tejasvi Ramadas Sagar
INTRODUCTION
Cannabis which is better known as
“Marijuana” is intended to use for psychological or medicinal use and it has
been proven to be a safe and effective treatment for the symptoms of many diseases such as AIDS, Arthritis, Alzheimer’s, migraines, hepatitis C, glaucoma,
psychological conditions and many more. Marijuana has impacted human lives in a rampant manner. But
lately, this has been legalized for recreational use. It falls under the
hallucination drug category. Initially Marijuana was used only for medicinal
purpose, but few states in the USA have legalized Marijuana for recreational
use.
In
Project Case study-1, two insights were proposed which are given below
1.
Prospective
states where marijuana can be legalized based on sales, quality and average
price per ounce.
2.
Legalization
of Marijuana has negligible impact on crime rate of Denver county.
In this project report, sales and crimes of Denver county in
Colorado with respect to the Marijuana is analyzed. The vote to legalize
Marijuana was passed in September 2012, but, it was not until 2014,
legalization was official. January 2014 is considered as the start of
legalization period.
The project will provide a deep dive into the overall crimes and
then drill-down to Marijuana crimes related to sales of Marijuana. The analysis
is performed with comparison of pre-legalization and post-legalization period
and how the behavior of crime is related. Our target audience are the
pro-weed(Marijuana) people who want to legalize Marijuana in more states for
business purposes. We would like to showcase that legalizing Marijuana for
recreational use does not affect the society in terms of crime.
Data Sources
Below are the datasets used to analyze the sales of marijuana in
Denver County before and after legalization, overall crime statistics of Denver
county and marijuana related crime statistics of Denver.
Marijuana Sales Dataset:
The primary data set which gives us the overall sales of Marijuana
in the US for the period of 2011, 2012 and 2013. This dataset was scraped from
PriceOfWeed.com using a Python script where the records of weed sold are
maintained and exhibited in GitHub folder for public use. The dataset contains
1.3 million sales transactions from all the states of USA.
Marijuana gross sales for
2014-2016 (Month wise totals):
Post
legalization (from 2014), County of Denver maintains record of sale, possession
and cultivation of Marijuana and its medical usage. The data is maintained by
County of Denver and is available for public use.
Denver overall crime dataset (2012
- 2016):
The secondary dataset includes criminal offenses in the City and
County of Denver from 2012 to 2016. The data is based on the National Incident
Based Reporting System (NIBRS) which includes all victims of person crimes and
all crimes within an incident. The dataset was obtained from County of Denver
website where all the crimes are recorded and the period from 2012 to 2016,
consist of both pre-legalization and post-legalization Marijuana period.
Denver Marijuana Crime Dataset
(2012 - 2016):
The secondary dataset includes criminal offenses related to
marijuana in the City and County of Denver from 2012 to 2016. Data in this file
are crimes reported to the Denver Police Department which, upon review, were
determined to have clear connection or relation to marijuana. The data is based
on the National Incident Based Reporting System (NIBRS) which includes all
victims of person crimes and all crimes within an incident. The dataset
contains 1254 records of reported crime in Denver county.
Understanding the Data
The project involves multiple data sets which gives information
about the sales of marijuana and crime statistics of Denver county. It is
important to understand the attributes and nature of the data to derive useful
insights. Below are some of the key information which will be helpful to
understand the data sets.
● The pre-legalization sales data
was from priceofWeed.com which is one of the channels which recorded the sale
of Marijuana. Post legalization period, County of Denver maintained records of
Marijuana sold for both recreational and medical use.
● The crime data set is from the
official website from County of Denver. The data set has categorized into
different crimes based on the offense type. Based on year/month, we can analyze
the trend present in the data.
● As the vote for legalization was
made official in 2014, the same will be considered as the start of the
post-legalization period.
● “Number of records” is a measure
generated by Tableau which is used to project the count of crimes committed.
Below are the important fields in each of our data set and its
description
|
Data set
|
Column Name
|
Description
|
|
Sales
data set
|
Month
|
Name
of the month
|
|
Sales
data set
|
Year
|
Year
the sale was made
|
|
Sales
data set
|
Gross
sales type
|
The
type of sale – Marijuana sold for medical or recreational use
|
|
Sales
data set
|
Gross sales
|
The
total sales month wise for the post- legalization period.
|
|
Crime
data set
|
Reported
date
|
The
date on which the crime was reported. To be used in Year/months format too to
visualize.
|
|
Crime
data set
|
Offense_Catgory_Id
|
Records
a broad category of crimes committed
|
|
Crime data set
|
Offense_Type
|
Reveals
the detail about the type of crime committed.
|
|
Crime
data set
|
isCrime
|
Represents
1, if it is a crime, else 0
|
|
Crime
data set
|
isTraffic
|
Represents
1, if it is a traffic accident, else 0
|
|
Crime
data set
|
First_occurance_date
|
The
date when the crime first occurred, could differ from the reported date.
|
Exploratory/Statistical Analysis
Understanding the dataset is more important and therefore, we need
to explore the data and perform analysis. Performing exploratory data
visualizations helps us to understand different patterns and trends in within
the data. This helps to make business decisions based on the insights
developed. Following are the exploratory/statistical analysis used in this
project:
· Box plots are created to analyze
the distribution of gross sales during the post-legalization period.
· Trend lines are generated to
infer, if there is an increase/decrease in crimes committed.
· Created calculated fields to
implement ratio of crimes committed per $1,000 sales, which gives relationship
between sales and crimes related to Marijuana.
· Small multiples technique is used
to perform drill-down analysis about the percentage change in crime committed
between 2012-2016.
· Difference in numbers of
drug-marijuana crimes is compared for post legalization period.
Insights
Based on the analysis performed, we have derived 3 insights and
used different visual encodings to help audience understand the visualization.
Following are the insights and encoding used.
Insight 1: “Traffic violations are
fairly constant whereas, overall crimes are increasing”
Observation: From this viz, traffic incidents
seems to be fairly constant but, there is a rise in all other crimes from 2012
to 2016. The trend line depicts the same.
Visual encoding: This is a bar chart which helps
to analyze the total crimes committed across years. Color encoding depicts
different years and classified by traffic violations and crimes
Insight 2: “Sales peak from late
Summer to Fall”
Observation: This viz gives the distribution
of gross sales of Marijuana used for both Medical and recreational purpose with
help of box plot. Across the years, the sales is on the rise and attains peak sales during late summer to
fall. Also, the sales are lower during the beginning of each year.
Visual encoding: Box plot is used to understand
the distribution of data and inspect, if there is a pattern in marijuana sales.
Upper and Lower whisker values are used to infer the variation in marijuana
sales against each month after legalization.
Insight 3: “Crimes related to
marijuana have dropped due to legalization”
Observation: As traffic violations are fairly
constant, categories of other crimes needs to be analyzed. Applying small
multiples, the graph displays a drill-down analysis and percentage difference
of each categorical crimes. Different categories of crimes have its own trends,
but focus is on drug-alcohol category since marijuana related crimes under
drug-alcohol category. As per the visualization, drug-alcohol has been
increasing but at decreasing rate.
Visual encoding: Color encoding is used to
differentiate year and labelling helps us to understand the percent
increase/decrease in a specific category of most recent year. Small multiple helps
us to portray all the categories in a single pane.
Observation: Marijuana related crimes are
categorized into 3 types - cultivation, possession and sell. There is a
declining trend in possession, and there is no sharp increase/decrease in sales
& cultivation.
Visual encoding: Bar graph gives a clear picture
about the difference of crime committed between previous year and current year.
Color is encoded for year. The charts are organized by types of marijuana
related crime committed.
Observation: The dashboard is used to compare
crimes committed per $1,000 sales during pre-legalization and post-legalization
period. During the pre-legalization period, there is a spike in crimes
committed and then during the end of the pre-legalization era, its fairly
constant. Whereas, there is a decrease in crimes committed, since the begin of
legalization, despite increase in sales. This portrays that Marijuana
legalization has minimal effect on Marijuana crime committed.
Visual encoding: Dual axis plotting is used to
depict the sales and crimes committed as a bar and line graph. Color encoding
is used on sales bars to differentiate the legalization period (orange-
pre-legalization and green as post legalization). Dual axis is used to compare
the two characters - sales and crime in a single frame.
Story
This story tells us how the characters - sales and crimes affect
Marijuana legalization in Denver, Colorado. Colorado voted in favor of
legalizing Marijuana for recreational use in 2012, but the law was made
official in 2014. The story is narrated by analyzing how our characters
influence the pre-legalization and post-legalization era. This story uses Logical
Rhetoric for the impacts on legalization of Marijuana. This story is based on the 3-act play model
where we have a beginning, middle and an end. This story model has created a
tension among the audience and resolves it by stating facts and guiding them
towards our insights. Following is the story portrayed.
“Impacts of Marijuana
legalization”
The story has analyzed the data related to marijuana, and narrate
how marijuana sales and Marijuana crimes impact the decision of legalizing marijuana
for recreational use.
Setting
Place: Denver, Colorado
Time: 2012 - 2016
Characters: Sales and Crime
· The story begins by comparing sales and crimes
committed/$1000 of sales during the pre-legalization period. Here we infer that
after a sharp spike, the crimes committed is constant till the end of 2013.
· Marijuana is used for both medical
and recreational purpose. The gross sales, during the post-legalization period,
are increasing since 2014 - when Marijuana became legal. As observed, the sales
are highest during late summer to Fall.
· It shows that the crimes committed
is on the rise since 2012 till 2016. The challenge is to find whether it was a
bad decision to legalize Marijuana for recreational purpose.
· With the help of small multiples,
the visualization explains the changes in categories of crimes committed. Since
we are focused on Marijuana related crimes, so we observer the drug-alcohol
category of crimes committed. Drug-alcohol category is increasing at a reducing
rate annually. To further investigate, we now look at Marijuana crimes alone.
· As we examine the visualization,
there are three types of Marijuana related crimes- Marijuana-cultivation,
Marijuana-possess and Marijuana-sell. By evaluating the change in Marijuana
related crimes, it depicts a decline in the Marijuana-possession and a constant
Marijuana-sell & Marijuana-cultivation crime. This shows that Marijuana
crime is not on rise even though overall crime is increasing.
Story flow
Conclusion:
The comparison of crime rate of pre-legalization and post-legalization
in Denver with respect to sales shows that Marijuana related crime is
decreasing after legalization of Marijuana, although, the sales are on the
rise. Finally, Crime related to marijuana is decreasing when marijuana is
legalized.
Comments
Post a Comment