What Are Data Visualisations?
Data visualisations play a key part in understanding your business or your customers. Essentially, they’re images of stored data you’ve collected over time. But they can hold secrets you’d never really see if it was constructed the way you need it to be.
There are literally hundreds of different types of graphs for displaying data. I’ll go through some of the best tools for data visualisations.
#Five: Google Sheets
Google Sheets is something we all know about, it’s the free and in some ways, better version of Excel. I spent years disliking Excel until I started learning how to use and combine formulas to get the exact data I needed. It’s not long been confirmed that Excel is now considered a programming language.
I digress though, Google Sheets makes it exceptionally easy to make a graph. It’s literally a case of highlighting the data you need and clicking “Insert Chart” and boom, you’re in. Some of the initial graphs you make just look cheap, and quickly made, which it is, but the options available to customise them further and combine other data sections with things like slicers and pivot tables really make this tool something that can make some fantastic dashboards for dynamic data visualisations.
# Four: Power BI
Power BI is a Business Intelligence data visualisation tool that lets users create dashboards for their clients or internal teams. Power BI is a fantastic suite of tools by Microsoft and the level of customisation they offer is, without doubt, something special.
They’ve created their own programming language similar to the one used in Excel called DAX to make custom columns based on calculations to drill down on your data even further and make better decisions based on the data you’ve collected.
Even though this is a stock image the premise is still the same. Just by looking at the displayed data Managers, CEO’s and the higher-ups can all see how effective their strategies have been and even identify specific areas to target.
Having an effective dashboard is useful. But making it look amazing is something that brings me on to the next item on my list.
#Three: Tableau Public
Tableau has a great community behind it which makes it one of the top data visualisation Softwares packages out there. You can create dashboards online, or use a desktop version. Tableau has free versions that you can create and publish your vizzes to the web or with the enterprise version you can create servers that then allow you to create and send dashboards to the people you want to safely and securely.
# TWO: MatplotLib
This is probably one of my favourite visualisation libraries. MatplotLib works with Python and it’s so highly customisable that you can create some really interesting data visualisations with it. The guardian uses the library to create some publication-worthy images.
Matplotlib has an extremely basic initial output. You can create a graph with just one line of code (If you’re using it in jupyter notebooks of course) and you can have something in front of you within seconds. Looking at this graph above though it would have taken much more tweaking to get it right. But the Guardian uses images like this all the time and even though it is just a line graph at the end of the day. Its simplicity is what this is a nice piece of content to look at for data storytelling.
While I’m praising Matplotlib, check out a post I did for LinkedIn.
Data Visualisations For InfoGraphics
Firstly, if you’d like to check me out on LinkedIn here’s a not so subtle link to the actual post itself where we may connect on a business level where you can get some insights into more data analytics and data visualisations I like to put together. But for those of you that can’t be bothered to do that, here’s the image I created and solely made in Matplotlib. I did however use 2 lines of code from another library for data visualisations called Seaborn, which I’ll talk about shortly.
Again, this entire thing used something like 300 lines of code to create. There are actually 6 different graph spaces on the image and I’ve customised each one in order to create something unique and interesting. Everything was customised, the line widths of the average episode ratings, the colours even a few extra lines importing new texts. It was a labour of love, and also pain, but perseverance got me there in the end.
Seaborn is another library that works on top of Matplotlib but it creates some really stunning data visualisations right out of the box. Used some seaborn data visualisations to show off how data was dispersed when conducting the worlds largest keyword study. I used one of their Violin plots to show how areas of keyword intent were concentrated in different areas.
Seaborn uses some really nice colours and the customisation options are still quite varied so you can make the perfect data visualisations. Even if it’s not apparent right away stack overflow will have plenty of information to help you. Although for some reason the people that answer the questions almost always use loops to which is confusing for beginners.
You can use any of the tools available, personally as a coder (not a programmer) I love using Python to make my images such as Matplotlib and seaborn to create data visualisations. There are so many different tools out there that make some really amazing images but each tool is different and the learning curve for each one is different too depending on your current experience. I’ll leave a final bonus tool that has the ability to really stretch the limits of what’s possible and will be on my list of stuff to get a handle on for 2022.
Check out Visual Capitalist to see some really amazing data visualisations that I’m currently dreaming of creating. According to their FAQ page, they use Adobe Creative Cloud, particularly Adobe Illustrator to create their infographics.
Check out more stuff from me on our company blog page. I did something pretty cool regarding the frequency of numbers appearing in natural life called Benfords Law, Google Easter Eggs, and also a warning post about how correlation does not equal causation which may be an eye-opener for some of you and does have some useful tips about how to treat data when you find something that looks like a pattern.