Oil prices and its relationship with the amount of people drive in USA

Oil Curve

Atmajitsinh Gohil

Wednesday, January 14, 2015

I always try to replicate new york times visualization. Eventhough replicating them exactly is not possible but i reach pretty close to the actual image using R.

This plot was generated using the oil prices and the distance driven at New York Times.

The image shown above was generated using R and further updates were made using the Inkspace. Once we generate the plot we import it using a pdf format and enhance it by adding labels in Inkspace.The data is available here . The readers should save the file in a folder as oils.csv.To replicate simply copy paste the entire R code mentioned below and you should get a very similar looking plot. Before running the R code make sure you reference the folder that oils.csv is saved in R. R will always search for its file in its current directory.

setwd("C:/Users/agohil/Book")
embargo=read.csv("oils.csv")
par(mar = c(5,7,5,7))
plot(embargo$percapita,embargo$price, type ="n",axes = FALSE,xlab = NA,ylab =NA)
abline(v= c(7000,8000,9000), h = c(25,50,75,100),lty = 3,lwd = 1.5)
lines(embargo$percapita,embargo$price, lwd = 2, col = "#28363F")
points(embargo$percapita,embargo$price, lwd = 2, pch = 21, bg = "white", col ="#28363F", cex = 1.5)
axis(1,at = c(7000,8000,9000),tick = FALSE, cex.axis=0.5)
axis(2,at = c(25,50,75,100),tick = FALSE, las= 2,cex.axis=0.5)
axis(3,at = c(7000,8000,9000),tick = FALSE,cex.axis=0.5)
axis(4,at = c(25,50,75,100),tick = FALSE, las= 2,cex.axis=0.5)

To import the data :

embargo=read.csv("oils.csv") # to learn more about the function type ?read.csv in R console window

Now if readers wish to observe the data partially they can do this by using the head() function to view the first 6 entries or tail() to view the last 6 entries of the data.

head(embargo)
head(embargo, 2)
tail(embargo)

Now, we are ready to plot our image in R. Note that no image will be generated here. The argument type = “n” will supress the plot. So why do we do this? Well we would like to customize the entire plot with axis and gridlines. The first two arguments in the plot are X (distance driven by individuals in USA) and y (price of a barell of oil). We will also supress axes and labels by using the axes = NA, xlab = NA, and ylab = NA respectively.

par(mar = c(5,7,5,7))
plot(embargo$percapita,embargo$price, type ="n",axes = FALSE,xlab = NA,ylab =NA)

The par() function is uses many different arguments and we have used mar=c() to specify the margins of our plot. Readers can use type ?par to get more information on the function. The lines() will generate a partial plot.The abline() function will plot the gridlines.

abline(v= c(7000,8000,9000), h = c(25,50,75,100),lty = 3,lwd = 1.5)
lines(embargo$percapita,embargo$price, lwd = 2, col = "#28363F")

But, we like to generate the line and plot the points on it. We can do this by plotting the lines() before the points() function - using the following lines of code:

points(embargo$percapita,embargo$price, lwd = 2, pch = 21, bg = "white", col ="#28363F", cex = 1.5)

We can now generate the axis and label all the 4 sides using the following lines:

axis(1,at = c(7000,8000,9000),tick = FALSE, cex.axis=0.5)
axis(2,at = c(25,50,75,100),tick = FALSE, las= 2,cex.axis=0.5)
axis(3,at = c(7000,8000,9000),tick = FALSE,cex.axis=0.5)
axis(4,at = c(25,50,75,100),tick = FALSE, las= 2,cex.axis=0.5)

Prevalence of Obesity in USA - 2012

Objective: The main objective of this map is to study the prevalence of obesity in USA.

Critique: Before any of the experts in data science or data visualization criticize my project in their blogs let me inform you that i am trying to learn and use tile mill, Qgis and scape toad via this data. I am quiet aware that

1) Alaska looks bigger than it should be. I am not sure why but the shape file is the same

2) A better way to show the prevalence of obesity is via a choropleth map over time but www.cdc.gov  already does this for their viewer.

3) A better way would be to show the same image for 2 different time span and show them side by side but the data for obesity in 1990 is not available for all the states and so it would not be a fair comparision.

usefulness of the distorted map: A good way to use a distorted map like this one is to show population or real estate prices in each state, or internet users. The possibilities are endless.

Technology: I have used a QGIS (open source) for creating shapefiles, Scape toad for creating distorted map and tile mill for making the map look prettier.

Refferences:

1)www.cdc.gov - for data

2)https://www.census.gov/  - for USA states shape file

3)http://spatial.ly/2013/06/r_activity/ - for inspiration

Walmart spread in USA - intro 2

Idea: To use Tile Mill to show the location of Walmart in USA.

Limitations: The data for the project was downloaded from Flowing Data . However, i cannot provide you with the data for this project as the data is not freely available. You can download the data from Agg data at a price. The cheaper and better alternative is to subscribe to flowing data tutorials and download the data from the flowing data itself. If possible i will try to download some geocoded data and provide it to the reader but for now you will have to find your way out of this situation.

Step 1: Download

Tile Mill - its an open source project. You can also browse the documentation section, which are very well written. 

Step 2: Data 

The data file is a .csv file with 7 columns.In order to plot the data in tile mill you would need to geocode the data. The data should contain columns lattitude and longitude, else you will get an error.

Step 3 : Power up your Tile Mill

An important note to keep in mind is that by default mapbox will create a folder under your user /documents/mapbox/projects/<name of the project>. Every time you create a new project it will create a new folder under mapbox /project. Hence in case you want to ad images or svg you need to save them under the project folder else you will not be able to use them in Tile Mill.

Open tile mill and click on new project enter a file name and click add, tile mill create a folder on screen double click and open the project. You see that tile mill will use default country data and launch a world map. Users do have the flexibility to change the map but for now lets keep it as it is.

The main screen is divided into 3 parts. The left most part is a menu which can also be used to add data(layers), labels, legends etc. The center contains the world map along with the data once you load it. The right most part contains the css sheet which is mainly used to edit style sheets. 

Step 4 : Upload Data

to upload the data , save the data in a csv format in a drive or desktop. Click on the last icon on the left most part of tile mill(image for reference). Tile mill will now open a new window. Now enter the ID click browse followed by save and style. The save option will only save the data but save and style will plot the data as well.

You would now be able to see the data points plotted on the map. As you zoom you will be able to observe the data at granular level.

Step 5: Customizing 

In order for the map to look like the image on the very top you will have to add different colors. You can easily do this by selecting the colors from palette at the bottom of the right most part of tile mill.As you select different colors, and save, for the background, borders and maps you will observe that the changes will simultaneously take place on the map itself. The CSS code will also change. If you know how to code you can make changes directly in the code window.

Step 6: Export 

In order to export the data click on the export drop down and select the format in which you would like to export. I usually use png or pdf. Now tile mill will allow you to edit the map and select the area of the map that you would like to export.Sometimes you will have issues selecting the area, i dont have a solution for that.

If you go back to the same export drop down you will observe view exports this will then allow you to save it once the saving process is complete.Tile Mill will also tell you where it has stored the image. Lastly, you can open an account on mapbox and upload your maps by exporting the format MBTiles. I have never used xml but you can explore that option as well.



  

Inequality and Wealth

Motivation:

The main idea of this article is to introduce readers to Google Ngrams.

Introduction:

Since Obama mentioned the word inequality in one of his lengthy speeches - every newspaper, columnist and every economist i follow had one or two articles on inequality, wage gaps or measurement of inequality.So i thought why not write one myself.

But i had something different in mind i wanted to use the power of big data along with the algorithm of google so i deployed "Ngram". If you have never used Ngrams then this is the right time for it. Just google Ngrams. Ngram basically searches google books huge database to come up with an line chart that shows how often the word has been used since 1800 to 2008.

The best way to interpret this graph is to observe the use of word inequality and wealth since 1800 in books scanned by google. After the second world war race between inequality and luxury  started  getting interesting and it seems like word inequality has gained momentum since 1980. So it is not just an Obama speech that has caused the rise in study of inequality but it has been popping its ugly head since 2000.

Fugly Map - intro 1

Inspiration: I am not that smart to come with an original idea like this one. I came across a tweet by Professor that said that he had asked his students to come up with an ugly USA map and he tweeted the image. I wanted to do something very similar. The link to the professor site : http://www.brenthecht.com/

Tile Mill: The image was created using tile mill an open source project by Mapbox. Its free to download and the documentation is very well written. Its a great tool to quickly generate graphs. It took me like 2 hrs to come up with the image posted above. This image is very similar to what the professor had tweeted.

I will try to write a small introduction on step by step instructions on creating this UGLY map. But if you are curious to learn may be this link will help you get started .

Off course similar things can be done using R. Somehow i think Tile mill is easy to learn and very handy. If you like to play with maps, data and not great at coding then this is a great tool for you.

Inequality :1

Inequality :1