FreeSight is a “small data” management tool (up to 10 million rows of data by tens of thousands of columns), a business process development engine, a spreadsheet, an ad hoc query tool, a small data business intelligence tool, a data analysis tool, a data cleansing, integration and transformation tool, and a report generator – all wrapped into one powerful and easy-to-use tool for anyone who uses PCs and spreadsheets for data manipulation.
FreeSight was designed to enable you to change your approach to data management and analysis – allowing you flexibility to experiment with your data without worrying about messing up or losing any work you want to save.
To work productively with FreeSight you need to understand three things:
- What is the Home Page View and what can you do there?
- What are the various Data Tables and Analysis Tables and how do they connect?
- Creating your FreeSight model and understanding the data flow
I. Home Page View:
When you open FreeSight you come to the Home Page View, a blank workspace that is a working area for you to play with your data. This is where you import your data and set up the workflow for the operations you wish to perform. From here you are able to see all of the steps your data goes through as you complete your work.
You can import as many datasets or files as you need. You can link data objects together, or unlink them. You can make copies of datasets to experiment with different filters, views, charts and analyses.
The operations you create and link together on your Home Page are the building blocks for completing just about any type data manipulation task you can imagine. As you work, and when you are finished, the Home Page View is the complete illustrated workflow for your project – in effect, an audit trail of your work.
By comparison, most spreadsheet-driven applications are like black box operations to all but the most expert users. Data is fed in on one side, and through all types of macros, calculations, lookups and formulas that are often hidden behind the scenes and exceedingly difficult to reverse engineer, data comes out the other side as results and reports. FreeSight changes this picture, literally and figuratively.
The Home Page View, like an audit trail, shows you every operation you’ve completed in your work. You can look into the details of each operation by opening up the various links and tables displayed. Everything is transparent.
II. Data and Analysis Tables:
The building blocks of a FreeSight model are the various Data and Analysis tables, which are where you complete much of your work. Complete Table descriptions are in FreeSight’s extensive Help sections on the website and also within the FreeSight product.Data Tables (e.g. Import, Manual or Append Table) have many similarities to a standard spreadsheet in that you can edit or correct your data, add or delete rows and columns, manipulate data with formulas, filter, sort, etc. But you can do much more here as there are many functions and features that are unique to FreeSight. (These functions are also listed in FreeSight Help.)
For example, unlike a spreadsheet, every edit or change you make to your imported data is clearly identified. Every added row or column, every edited cell value, and even column or field name changes are highlighted, with the original value displayed when you “hover” over a modified cell. Your original source data is NEVER altered.
Other unique features available to you in a Data Table include: the ability to retrieve data you deleted from your table, even after you’ve saved and closed your file; the ability to make global changes to specific values (correct matching values or remove correction commands), powerful sorting and filtering options, non-sequential undo or redo commands, drilling down to source data (view source data), the ability to “reset” your data to its original row or column order (reset row/column order), and more.
Analysis Tables are used for data summarization, analysis and charting. They look like and function much like Excel pivot tables, but are easy to learn and very flexible. Powerful auto-charting features enable you to immediately see visualizations, charts and summarizations of your data that you might not have even thought to explore. Or, like a pivot table, you can drag and drop your data to display the fields and summarizations as you wish.
Unlike pivot tables, Analysis Tables also allow you to perform the same types of operations that you might perform in a Data Table, such as making corrections, adding columns, formulas and more – for purposes of summarization, charting and reporting.Plus, dynamically while you work, FreeSight automatically creates charts and graphic displays of the changing views of your data.
More powerful features of Analysis Tables are discussed in the Data Analysis write up.Reference Tables are an easy to learn and efficient way to manage data such as product lists, customer lists, employee names, etc. – any repeating data used for reference purposes. Because they are so easy to create, manage, and are automatically accessed by FreeSight wherever you wish to apply them, they completely replace the need for VLookups or any other equivalent mechanism to access related data stored in a separate file.
You can create Reference Tables manually, or FreeSight can create Reference Tables for you automatically (Auto-Create Reference Tables) by extracting the sets of related reference data from imported files, and creating individual Reference Tables. You can manage these tables easily, explained in more detail in the Data Cleanup write up. You can also visit FreeSight Help for more information or our three-minute Video Tour of this feature.
III. Creating Your Model:
Creating your model is simply following the sequence of operations you need to perform with your data to deliver the results of your analysis – using FreeSight Tables and connecting links to complete each step in your process.
FreeSight operates like any IT input-process-output system. Data is imported, joined, cleaned and normalized, and then analyzed. Finally, results of the analyses are exported into whichever reporting mechanism you wish to employ. We frequently use this graphic to help visualize a standard process flow.
You can visibly see your complete process and workflow on your FreeSight Home Page View.
When building models, you can create as many Tables of any type as required to perform all the operations you require. As you create and work with the different Tables, you’ll find that they are connected by arrows and links, illustrating the flow of your data. You can make edits, corrections and changes to any Table, at any location within the model. Since the data flows in the direction of the arrows, any edits or changes you make are automatically applied downstream (but not upstream) in the flow.
As you’ll see in an example below, FreeSight contains built in controls for transparency and audit trail. The workflow and the Tables within show every action you have taken. FreeSight never actually changes your original source data, yet gives you complete ability to edit and correct errors. Because you can create as many Tables as you wish, you have unlimited capability to experiment and play with data without impacting the charts, graphs and tables that you’ve already created and wish to save.
Finally, you only need to do this work once. For recurring tasks, when your source data is updated or refreshed FreeSight updates all of your analyses and reports automatically.
In the example shown below, a typical FreeSight model, you’ll see Data Tables on the left side of the Home Page View that contain copies of all the source data that was imported: Sales Data for three years, a Customer Table, some Product Data, and some Region and Sales Rep Data. In a live model, when you open any of these Tables you can see and work with the data itself.
Following the flow of the arrows, you’ll see that the three Sales Data Tables are appended together to create a 3 Year Combined Sales Results Table. Concurrently, the Product Data and the Region and Rep Table have had relevant data extracted into Reference Tables.
Some manually entered data (Mosaic and SIC codes) have been joined to the Customer Table, which has also become a Reference Table. These Tables, in turn, are also joined to the appended 3 Year Combined Sales Results Table. (The joining, by the way, took seconds to complete.) This Table is now ready for analysis.
We wish to create a few different sets of analyses: By Region, By Sales Rep, and By Product, and Totals – so we have elected to create four paths of analysis. Following the data flow, you’ll see three linked-copies of the 3 Year Combined Results Table, each filtered and using only the data subsets required to analyze results by Region, Sales Rep, and Product. Because it is so easy to do, we can create and look at dozens of charts and views of the data. In the end, the seven most descriptive charts and tables are exported directly into a presentation template for distribution.Creating this model, including importing and joining all of the data tables, and then creating analyses and reports, would have taken a user a fraction of the time it would have taken to complete this work using most tools. Better still – when new data arrives, all of the tables and reports are updated automatically simply by dropping new data into the model.