Do you ever come across a graph like the one on the right and think I want to include that data in my work?
If so, this post will show you a simple tool to digitize scientific plots.
The post will take you through the steps of setting up the plot, defining the points, and capturing the data. Note that permission is required from the journal publisher to republish the data, so be sure to do this before using it in a journal article, presentation, or poster.
FindGraph is a simple and easy-to-use tool for scientists, engineers, and graduate students to digitize scientific plots. This tutorial is completed using a 30 day demo version downloaded from their website. Note that I have no affiliation with the company, and am just providing this information as I have found it useful in my work.
How to Digitize Scientific Plots
The first thing to do is to order findGraph or download the free demo version here. I am using the evaluation version for this post (Version Number 2.622). One major drawback with this tool is that the install only works on Windows computers.
Step 1: Workspace Setup
To start the process, open the findGraph tool. You will see an entry screen that looks like the picture below. Note that a sample browser generally opens as well. You can explore this for more information on the tool, or close it and continue on with the tutorial here.
I have found success using a square workspace, as the image is not skewed when imported. To do this right click on the white space and select “Plot Details”. Under the “Page” tab, set the “Width” and “Heigh” to be equal. On the very small laptop I was using (the only computer I own now that is not Mac or Linux based), this was 12×12. However, you should be able to fit a larger workspace to your screen if needed.
Step 2: Import Scientific Plot
In order to digitize the graph, you need an image of it. The best images come from screen grabs of PDFs, however images from your phone work as well. To import the image click “File” then “Open Image”. Select your image and it will load into the workspace.
Sometimes the image is skewed or rotated. FindGraph includes tools to fix this which can be found under “Plot” then “Picture”. For example, images that are rotated can be moved back to straight up-and-down under the “Placement” tab.
Step 3: Select and Digitize Data
The process for selecting the data is quite easy. FindGraph includes many automated tools for finding and manipulating the data; However, I have found the manual process is easy enough to use, if you know the steps.
Start by clicking “Data” and “Add from Picture”. This gives you a digitize pop-up window, that steps through the process. Click the second radio button and click next.
Select the plot reference points. I generally only use three points: the intercept of the axes, the maximum x-value tick, and maximum y-value tick. This seems to give reasonable accuracy when extracting the data. Hover your mouse over the points and use “i” on the keyboard to select.
Once you have selected the points, enter their coordinate values in the window for “Step 3”. Step 4 and 5 allow the data to be automatically detected; However, I do not use them. I click next on Step 4. On Step 5 I select the top 3 or 4 colours and click digitize, moving onto the next step.
On step 6, click the button “Remove all points” and add your data points manually. Again use “i” to add the points. The “d” key can also be used to erase points if needed. In the picture below, I selected all of the circle data points (click on the image to enlarge).
The last step is to export the data. The last window will show the symbols you selected. Click “Finish” to exit the “Digitizing Scientific Plots” window.
To save the data click on “Data” and “save points as”. This will give you a summary of the points and allow you to set the file name. I like to check a few of the points in the window as well as the final file, to make sure that the process worked correctly.
There you have it, a simple and easy way to digitize scientific plots and graphs! I have found it less troublesome to completely close the program and start from scratch for each dataset entered. I am sure there is a better way to do this, but this seems to get the job done.
Do you have any other tools that you like to use to digitize scientific plots? If so or if you have any questions, leave them in the comments box below.
I look forward to hearing from you and until next time, Happy Researching!