console.log("script ran")

dataviz tools / LEARN

Dataviz tools – Part 1.

2018-12-20

Tags: , , , , ,

I asked my Dataviz team to choose and discover a new data visualization tool what they have not used before. This post tells you what we have explored. Enjoy it and recommend us new tools.


Flourish

Judit Bekker

I choose to explore Flourish, a tool that is more suitable for data journalists than analysts with its easy to put together, but eye-candy ready-made charts. Even though it’s easy to use and looks great, – as expected – it only allows for a little flexibility.

Flourish offers 3 versions – no surprise, the same as all other BI softwares –, public where everything you do is visible for all the others, private (£39 / month) and business. The latter provides you with a bigger playground, you can upload custom templates, collaborate with your team and amongst other features it lets you remove branding.

The basic visuals are categorized by type (line, projection, scatter etc.), and there are some advanced charts available – that would take a lifetime to built in the well-known BI tools – like the chord, parliament, horserace or Sankey diagrams. They can be somewhat customized but has more limitations than flexibility: color scales are predefined (beautiful though) and fonts cannot be altered. On the positive side it works well with pictures and icons, plus super easy to import your data with an almost self-explanatory input interface.

My visualization:

To test it out, I put together a custom map for my yearly Prague trip with Flourish’s new marker map visual. If you click on the icons a short description of the place pops up along with a picture (if available).

Map: https://public.flourish.studio/visualisation/178107/


Charticulator

Mátyás Sereg

Charticulator is a Microsoft Research Project created by seven people listed on the website. The aim of the project was to create a flexible visualization tool that can create highly customized charts, that are not present in popular data visualization software or are only possible with using scripting languages or libraries like D3.Js.

Charticulator does not offer visual templates like major industry leaders (bar, line, etc.) but also does not require programming knowledge to build visuals. Even though it does not require programming knowledge the tool itself is built by using a number of Javascript libraries. The charts built in Charticulator can be exported as image files but most importantly also as visual templates that other tools might be able to use (e.g.: Power BI).

Pros:

  • It is an entirely online tool, so you don’t have to download any apps of programs, also no Chrome add-ons or anything like that.
  • Although it takes some time to get used to the interface it is fairly easy to use after you get the hang of it. Pro tip: even if you consider yourself an expert in other visualization tools, you can save yourself a lot of time if you read the Getting Started manual and watch the tutorial videos – I thought myself to be better than that…I really shouldn’t have 😊
  • Importing data is really easy, it is literally 3 clicks (so is replacing the dataset you are using).
  • After you have your data imported, you can basically start building the visual you want by selecting the types of marks you want on your canvas and populating them with your data by moving the columns to the appropriate places (x and y axes, colors, width, height etc. depending on the mark type).
  • Once you have the foundations of your chart the canvas you can start customizing it – and that is where Charticulator is a really handy tool. There are a ton of options that you can manipulate and fine tune to your needs. You can basically select all parts (they call it layers) of the visual – be it the background, the axes, the titles, marks, links… – you can change their colors, styles, size, type (according to what you are editing).
  • There is usually always a new layer that you can add to change and customize your charts or to give it a new spin you haven’t considered yet (scaffolds for example).

Cons:

  • Charticulator only works with CSV files and there is no option to edit your data in any way. This means that you need to very thoroughly prepare the data you want to work with.
  • There are certain chart types that need extra data preparation – like the attached Sankey chart. Here you need two tables, one that contains the nodes and the values and another that specifies the links between the nodes (child-parent or other relations).

  • There are no map charts – or at least I haven’t found the option to create one.
  • I found performance somewhat poor when working with bigger datasets, so you should limit the size of your data to a couple of hundred rows max (as per my tests).
  • There is no option to create interactive visuals or dashboards.
  • Even though I could create a custom visual type that I exported, I could not make it work in Power BI no matter how I tried. I assume that is a limitation of PBI that sometimes there are simply too many marks up on the chart that have no equivalents in Power BI that it is not possible to replicate the same visual output – but I might be wrong here. Even if it is possible to do what I tried, it is not trivial to do so, that is for sure.
  • Of course, it is not an enterprise tool, so saving your work, automating it or setting up security is not something you should look for.

In summary, Charticulator is a very good tool if you want to create bespoke charts that you want to publish or want to reuse in other tools, but you don’t have the necessary scripting expertise. It is more of a visual experimenting tool than a reporting tool of course. So, it is excellent for fooling around with your data in it and explore new visualization methods and chart types to find the one that best represents what you want to show from any given dataset. Once you get familiar with the interface and the options, it is fairly quick to work with it, but you should be prepared to move your data around quite a bit while doing so. The real expertise needed to use Charticulator is not how the interface work, but to know which chart type require what data structure – you get that, and you can build mostly anything in it.


Google Data Studio

Julia Borsi

Data Studio is Google’s free data visualization product for easy creation of reports and dashboards. I used it for the first time with the most recent Makeover Monday data set and here’s what I found:

Pros:

  • Pricing: free
  • User interface: easy to use, well structured and understandable
  • Collaboration: team members can collaborate on a dashboard the same way they would do with a Google document, which I found really handy. (
  • Data sources: besides uploading your own files, it can connect to any data available from Google. That includes Google Analytics, AdWords, Search Console, BigQuery, and more.
  • Customization: wide range of visualization options, full custom design and style controls. Users can also embed: URLs, texts or images to create visually appealing dashboards.
  • Report gallery: users can find inspiration in the Data Studio report gallery or use dashboard templates
  • Basic chart types: Data Studio offers the following chart types: table, scorecard, time series, bar, pie, donut, geo map, line, area, scatter plot, pivot table, bullet chart.

Cons:

  • Deep dive not supported: I think it was not designed to do complex analysis, the calculated fields were quite hard to use. I would recommend doing the analysis in an Office Product, then visualizing it with Data Studio.

All in all, I would recommend this tool for creating eye-candy dashboards that needs to be shared with the whole team, but would not use it for complex data analysis. You can reach my viz here: Bus Safety Performance in London


RAWGraphs

Tamás Gáspár

You have the data, but takes a lot to make a fancy chart for a presentation, a leaflet or a website? Then RAWGraphs is something you should have a look on. It has a lot visually appealing chart types. Just take a look on their gallery: https://rawgraphs.io/gallery/.

RAWGraphs is a web base tool. It has no desktop client, no need to install anything.

To create a diagram:

  1. Copy or upload your data, it either can be to their website or to any CORS-enabled website (e.g. Dropbox).
    It worked quite fine, took me ~2 minutest to start working on a ~10K row xls (wifi with 20Mbps upload).
    You need to merge (join) your data before, as you can’t connect multiple files or excel tabs in the tool.
  2. The fun part starts after, you are welcomed by the available charts, 21 precisily plus 1 where you can use your D3.js chart.

As an experiment I started a sankey. (I had bad times when tried to create in my mostly used tool.)

  1. It could not been easier. Pick chart type, drag and drop the dimensions you need and a measure for sizing the belts. 30 seconds 🙂
  2. Every change is instant on this dataset, you put a color on it immediately appears.
    Scatter plot, where all the rows are visialized is slower, but still decent.
  3. You have limited customization option here, canvas size, node’s indicator width (why? 🙂 ), oppacity, colors, and sorting the dimension values by size or name.
  4. Exporting options are vector graphics image (svg), pixel graphics image (png) and data model (json).
    Unfortunately you can’t save your work to continue later.

RAWGraphs states itself as mid tool what is “…providing charts that are not easy to produce with other tools.”. And it is accurate. When you have the data, it is easy to make it visual, but if you want data level changes, you need to that in your data file. There is no option to create new calculation or apply any filters.


Piktochart

Tamás Varga

Piktochart is an online, free-to-use app with the possibility to create visually appealing dashboards, infographics, presentation and printables. The user interface is simple and well structured and the you’re able to store projects online. For each type of visual you can choose pre-made templates to start with or you can create your own template from scratch.

There’s a huge variety of things to use, sorted by themes:

  • shapes & icons
  • photos
  • fonts
  • color schemes
  • easily editable chart types, maps

You can also embed videos, upload photos and visuals (jpg, png, gif or svg format).

Creations can be downloaded and shared. In the Inspire Me section users can share their visuals and can get inspiration from other users’ works. There’s also a Piktochart Infographics blog where you can read articles and ideas in relevant topics such as education, design, user stories etc.

Pricing?

Freely available, premium features available in three possible variation as Lite, Pro and Pro Team.

 

 


Infogram

Katalin Szalay

Infogram is a web-based infograpic and chart maker tool. Reports and dashboards can be created by dragging and dropping components onto a blank page or by modifying an existing template, it doesn’t require any coding skill or database knowledge.There’s a basic free version but to access more sophisticated features and extend limitations there are different price categories on a monthly or yearly payment basis.

Pro: Simple, easy-to-use user interface. Quick layout design. The layout is really flexible it can be customized easily.

Contra: very limited data source connection options (only files).

I would recommend this tool for:

Storytelling and presentation purposes or visualizing insights coming from data analytics but only based on well-structured and prepared, stable dataset which doesn’t change much over time.

I would not recommend this tool for:

Deep dive data analysis, data preparation, complex visualization.


Datawrapper

Ivett Kovács

I follow Lisa Charlotte Roston’s twitter  – who is the blogger of Datawrapper – and her datawrapper charts fascinate me every time. This is an open source tool that was developed by data journalists.  If you read Guardian you must have met with this kind of charts.

Free/Paid version:

Datawrapper offers a range of paid subscriptions. The free version’s good for one user. For a small audience or to complement a presentation, this version could be a feasible option. However it is good for only 10,000 views per diagram, so if you have more clients, you will have to upgrade to a paid version.

Pros:

Creating a chart is so easy, no installation required. There are only 4 steps.

  1. Upload your data through CSV, Excel, import from Google Spreadsheet, link to an external dataset or just copy and paste.
  2. Check and Describe the information what you would like to display on a chart.
  3. Visualize your data with your own design on a selected chart type.
  4. Finally, you can publish or embed the built visualization.

Bonus is the Datawrapper River. This  is an amazing extension. It is a platform of visualizations users can reuse and share those charts and maps on their own website for free.

Cons:

You cannot build complex charts such as sankey. This is not necessarily a huge con however, as it is not easily readable by many people. For online media, this tool offers the best quick and easy way to visualize your data. It may crash on datasets larger than 1000 rows though (this is my experience).You cannot customize, export or print your chart at free version.

My first Datawrapper chart:

I would recommend this tool to:

Journalists or other content creators who work with strict deadlines and limited resources but want to build simple, clean charts quickly to support their own online articles.


So,what is your favourite?

Let us know if you have any questions or feedback for these tools. We’re always happy to hear from you.

by Starschema Dataviz Team

4 likes

Author

Ivett Kovács

I love taking datasets into beautiful and informative visualizations. Besides data, I am also very passionate about traveling, surfing, cooking and exploring the outdoors with my dog.

Your email address will not be published.