GraphUX

Inspiration

I am currently working in a Biochemistry lab here at the University of Georgia where I and my fellow labmates use commercial plotting software on a daily basis. This software is often convoluted, complicated, and most importantly expensive. The purpose of this application to to allow the users to visualize their generated data easy and produce publication quality images at a fraction of the cost of the other commerical products.

What it does

Currently the software allows the user to upload a csv file which is then plotted as a scatterplot and fit to a linear model. The program then plots that fit onto the scatterplot and produces a residual plot and summary statistics of the model fit to the data. The fit is reactive depending on the slider which allows the user to control the cutoff range for the linear model. In addition, the program simultaneously updates the plot of the fit, its residuals, and the summary stats as the user changes the residual cutoff.

How I built it

I built it using the shiny packages in R.

Challenges I ran into

The main challenges were in making the modules reactive to the residual slider. In addition, I started implementing a zoom feature on the main plot, and I unfortunately ran out of time, so it selects the points but does not zoom yet.

Accomplishments that I'm proud of

The reactivity of the program.

What I learned

I learned how to make R programs using the shiny package.

What's next for GraphUX

In the future, I will finish implementing the zoom feature on the main plot. In addition, I will add the ability to fit the uploaded plot to several different models including nonlinear regression and several others. I will also add the functionality to upload several csv files to do a global analysis. Additionally, I will incorporate the ability to manipulate the color and sizing of the plots and allow the user to download those along with the generated statistics of their fits for publication purposes.

Built With

• r