Original chart:
These features have been preserved:
- Data positioning: downward-trending scatterplot as shown. Weight on the x-axis and MPG on the y-axis.
- Scales: Note the scales do not start at 0.
- Axis ticks and labels: both axes are labeled and there are tick marks at 10, 20, 30, etcetera.
- Color mapping to Manufacturer.
- Size mapping to Weight.
- Opacity of circles set to 0.5 or 50%.
For the 5 charts below, there is an image of each chart in the img folder, and the code files in correspondingly named folders where applicable.
Excel is a spreadsheet program which facilitates creating charts from data. It's easy to create basic charts in Excel. It's a little harder to customize them. In order to make the size of points matter, I made a bubble chart rather than a scatter chart. To specify different colors for points I added 5 series to the chart. To make the axis ticks line up the way I wanted them rather than with the start of the axes, I made separate series to provide the tick labels and the ticks themselves. I also learned that saving to a .csv file does not save the Excel chart in it. Practice certainly does make faster - the first time I made the chart it took me almost 3 hours; the second time it took about 7 minutes.
My sources for this chart include the following: source for using different series for colors and source for creating axis ticks with a series.
Tableau is another chart-making software. Compared to Excel, using Tableau was incredibly enjoyable and intuitive. You add data; you drag and drop columns into the axes you want them on; you right click them to see relevant menus; you double click an axis to edit it (for example, to simply change the tickmarks...) or right-click it to format it. Size and color of the dots each have 'cards', where you can set what data they depend on and change their characteristics. One problem I had was that the bubble size differences didn't seem significant and I don't know how to change the scale, so I made another column "weight squared" and used that for bubble size.
For Tableau sources I used the Tableau home page and help pages.
R is a language used for statistical analysis. It has a ton of built-in functions to make dealing with data and graphing very easy. I used RStudio to write an R script to make the chart. I couldn't get the ggplot2 library to load (I think this might mean a problem with my R version or path), so I just made the chart using normal R functions like symbols(), title(), and plot().
For sources, I used this video as an intro to using R and making scatterplots, and these three sites to find out about transparency, tick mark settings and size, and axes.
Matplotlib is a python library that makes creating and manipulating charts easy. I also used the pandas library to load the csv. Making the chart with matplotlib was really easy, particularly because I looked at example code, which made it much faster and much easier than not looking at examples. (Also much faster than trying to make the Excel chart. This may be my new go-to.)
My sources for this were a demo of a scatterplot and changing colors in scatterplots .
d3 is a JavaScript library. The main trouble I had with this was accessing the .csv file. Not because I didn't heed the ubiquitous warnings to run a local server, but because I kept getting CORS errors, which took me a while to see (so the console is where the error messages go!) and fix. The fix was that I needed to change the settings on my server (or on the requests) to allow myself to access the file; so I changed the server settings (in the server.py file). The other issue was understanding when I had to call the axis tick function on the axes (before displaying them). I used example code for this one too, so again much faster than it would have been without examples.
I used this code to fix the no-CORS error, this code as my starter code, and this code to help me change the axis ticks.
- For technical achievements I'd say getting the tick marks in the right places on the Excel chart with the series was pretty cool. This indicates that you could pretty much put any kind of symbol or label anywhere on an Excel chart you want; you just have to set up a new series. If you had to add a lot of things, or commonly added similar things, you could probably set up equations to do this faster rather than manually for a given chart.
- I was also happy to learn that you can make the plot with normal R functions; I'd like to fix the ggplot2 download error though, and see what I can do with that library.
- Fixing the CORS error so I could access my own file was nice; I'd like to learn how to send a request that can get past that kind of thing.
- Not much here; I tried to keep the sizes distinguishable and the colors similar to the original chart (matplotlib is the farthest off). That's about it.