How to upload data into Explainum from local file using Dropbox

It is possible to upload data to Explainum chart just from local folder, and have it updated every day. Below, I will explain how to do this using as example my chart where just for fun I indicate my mood and explain changes in it.

To do something like this you will need cloud-based file-sharing service, e.g. Dropbox. Every registered on Dropbox user gets 2GB of space for free which is more than enough for our purpose. Once you have installed Dropbox on your computer -- notice that it has created folder named Public.

Create CSV-formatted file, using either Notepad or Excel. If you prefer Excel -- in Excel create a table on a sheet. This is how it looks in my case:

 
Then save it using Save As command as Comma Separated Values (CSV) file to the Dropbox's Public folder. When you save file into a Dropbox's folder, Dropbox automatically uploads it into its web-based cloud storage.

After saving I recommend to open it once again in Notepad to check date format, as Excel can have its own date formatting preferences.

When the CSV file is ready get the public link for the file -- right-click on the file and choose in the context menu Dropbox -> Copy Public Link.

Finally, the last step -- go to Explainum, create chart, add new series using "CSV file over HTTP" data source. Then just paste the copied public link and choose appropriate date format.  Don't forget to check Autoupdate checkbox to force Explainum get updates from your file every 24 hours.

Complete uploading by following instructions. Your chart is created. Now, when you add new rows in your local CSV file, your chart is automatically updated every 24 hours with this new rows. View it, comment it and share it with others.

Here is how my chart look like (it's interactive -- try playing with it):



Every day I add new row with date and value to my local CSV file, which then is automatically uploaded into Explainum and added to the chart.

Explainum is about to launch beta

We're busy with last preparations before starting our beta-testing. The beta is still pretty basic, however the core functionality is functioning. It includes:

• Line charts with comments tied to areas of datapoints
• Embeddable interactive widget that contains chart and comments feed (not real-time yet)
• 5 data sources with possibility of daily auto updating (besides web-form)
• CSV-formatted data entered through web-form
• CSV files over HTTP which makes possible to build charts from CSV files located in local folders using file-sharing services like Dropbox
• Yahoo Finance -- daily stock prices and market indices
• MSN Money -- daily stock prices, market indices and currency rates
• Google Analytics -- a few basic indicators: visitors, pageviews, etc. (no filtering possible yet)
• Wiki for help and documentation

We're accepting applications for participating in closed beta testing program. Just send us a few words about yourself on explainum(a)gmail.com .

PS. Yes, gmail mailbox doesn't look cool, but it does what it should :)

UPDATE: Registration is open now.

What is Explainum?

Briefly, Explainum is a service for creating web-charts with embedded comments. Unlike many other charting tools, Explainum allows linking comments to ranges of data points in charts, thus making it possible to easily explain specific data trends -- spikes, dips, etc. It's like Twitter for charts. See the magic on the page with a few sample widgets.

Benefit from this feature is more significant than it might seem. Let me explain it.

I've been working in data analysis and visualization area for more than 10 years and know very well that trend charts, so popular in business world, used to reflect key performance indicators (KPIs) that are influenced by many factors of a business environment -- e.g. current customer preferences and intentions, actions of competitors, legal acts, macroeconomic trends, etc. And even sometimes weather. If you prefer, these charts contain encoded information about influencing factors and events.

Obviously, the knowledge of these influencing factors and understanding the degree of their influence is crucial for making right decisions. Therefore, when looking at a chart we usually want to know the reasons behind trends, because knowing reasons gives control. Every spike or dip, or any other deviation from pattern has a reason behind it, and we want to know it. Actually, we want to decode the chart. However, explanations and interpretations of a chart often remain divided across emails, documents, reports, presentations, etc. In other words -- everywhere, but not in charts themselves, because popular tools used for data visualization poorly help communicate the knowledge derived from this data. Actually, all what these tools do is they just get numbers, draw fancy pictures and stop there. What do they offer for communicating meaning of the data visualization and collaboration around it? What do they do for decoding? Let's see.


Excel
The most ubiquitous data visualization tool ever. Explanations are made as side notes directly in spreadsheets or as notes in emails with attached spreadsheets. Will these notes be valid when the data updates? Probably, no. And what about search?

Enterprise Business Intelligence suites
The most powerful data visualization tools. Very smart and sophisticated systems with millions of men-years of R&D actually follow the same way -- they get data, draw picture and stop there. What about human interpretation of this data? Can you find all explanations why sales in Q4 LY were less then expected using any of BI suites ? No. Hmm, are they really decision support systems as they claim to be?

Charting libraries and hardcoded charts
The cheapest data viz tools and therefore -- very popular. Frequently used to plot charts in web-applications. These web applications themselves can offer various social features like forums or discussion threads. However how well these forums are integrated with charts? Not really.

Google Finance
To be honest, there were some attempts to create charts with some explanations linked to data points. One of them -- charts in Google Finance. It was very smart move -- to link news to data points, wasn't it? However a few questions arise:

• Who selects the news for charts? Does everybody think that this is the most proper selection?
• I want to know what that smart guy thinks about this trend -- how do I know?
• Events may influence much  longer than 1 day, period of influence may be delayed for several days/weeks/months -- how about this?
• How about explanation of trends in past? Often true reasons became known when time passed only
• If I see a comment to a dot on top of a spike -- is it related to the rise or to the fall? How do I know?

Explainum
So we came to the point, where it would be nice to have a new visual data collaboration tool. In our vision, the tool should allow users:

• Build live charts
• Easily communicate their knowledge about influencing factors behind data trends
• Search comments related to a data range and data ranges related to a comment
• Keep valuable knowledge concentrated in one place
• Accumulate history, day by day
• Mix public and personal data freely
• Mix data series from various sources of different kinds

That's why we decided to build Explainum. It may look raw, unfinished or unprofessional in its early stage, however this is the first data visualization tool designed to keep knowledge about influencing factors behind data trends ever, and it does what is does. Try it, make use of it and share with us your suggestions how it can do better what it was intended for -- decoding charts.