How Much of the USA Population Can Your State Feed?

Idea: 

After reviewing the data viz, I thought it would be better to explain the data as food that can be produced by each state.  I felt that adding in the states' food possibility also would give the viewer a size reference and outcome of the dataset.  

Data Viz:

Here is the link to the Tableau Page.

Insight:

The insight taken from this data was how much of the US population could be fed if a state was dedicated to using the land just for growing food.  The amount of food that could be used to feed the US population for a few years was very surprising.  Most states could grow a crop like corn and wheat, and provide enough food to feed the US population for more than a year.

Project:

This was a data set for Makeover Monday on Data.world.  The project was to makeover a data viz from Our World in Data.  The task was to improve the data viz.  This project was based on food and land size required to make 1 gram of protein.

Tools:

• Tableau Public
• Microsoft Excel
• Google Image Search
• data.world

Data:

The data given was from Our World In Data.  This came with 10 food type categories and the amount of land needed to produce 1 gram of protein.  A very simple and small dataset.

In order to add the state size data, I had to grab data from the US Census Bureau website to get the data on state land size.  The data came in a choice of including water and just land without water.  Since the data is about growing food just the land data was chosen.

The last piece was to find out the population of the United States.  Data from Google which came from the Census Bureau stated a total of 325.7 million people in the United States.

Data Cleaning:

The cleaning of the data was extremely easy.  The data of the amount of land used for food was already clean.  It had the type of food and the amount of land needed in meters squared (m²).

The state sizes need some simple cleaning.  The lands that are not considered a state were removed from the set.  The District of Columbia (Washington, D.C.) remained with the data set.

The data was imported into Tableau Public with a left join to the state size and the food data set.  In order to get them to join a "Year" criteria in the "State size" data set was added.  This allowed for the "Join" in Tableau Public by matching the "Year" for both data sets.  The year used was 2017 for all of the data sets.

The other part of cleaning was the conversion of square miles into square meters.  This was handled by using the following formula;

Meters ² = Miles ² X 2,590,000

This formula was used in Tableau as a "Created Calculated Field..." to the dataset.

With the food value in "Protein", the goal is to find out how much protein the average person needs a day.  This was men at 56 and women at 46.  The value of 20,000 grams of protein a year per person was used, which was easier math and in the middle of the men and women values.

The value of needed protein for a human was taken for a period of a year would result in feeding America in a year period.

Grams of Protein a State can Produce = Amount of Land Needed per Gram of Protein ÷ State Size in Meters ²

% US Population That Can Be Feed = (Protein Produced by State ÷ 20,000 Grams of Protein per Person per Year) ÷ 325,700,000

Process:

The layout of the data viz shows the state shape and the name on the left side.  The right side is the food with a graphic representation.  The values displayed are based on the percentage of the US population that can be fed if all the land in the state was used to grow food.

The original bar chart is located in the bottom of the visual to display the massive difference in land used for animal raising compared to the in-ground grown food.

The top has the state for the selector.  A drop down was used for aesthetic reasons.  The original idea was to use a slider, but the slider interface made it difficult to find a particular state.  The drop-down menu was able to provide this type of feature.