Data analysis for humans

Variables is a highly graphical and amazingly intuitive data analysis package. Explore your data, create sophisticated models and prepare stunning visualizations, all with your fingertips.

Visual model building

Variables introduces a great new, yet completely natural, way of building statistical models. You simply position the variables, connect them to define the dependence structure and get all the details about the model at your fingertips! Variables assists you by suggesting possible model configurations it can handle. And it even animates all the transitions when you make changes so that you can instantly see the impact.

Great tools for any job

Some properties of the data are evident from a single variable. You can rely on descriptive statistics to summarize it, or employ a modern technique like kernel density estimation. And when you need to delve deeper into the structure of the data and analyze the interdependencies among variables, there's everything from simple correlations and scatterplots to more advanced generalized linear models.

Stunning visualizations

The old saying that a picture is worth a thousand words is as true as ever when it comes to data visualization. Why dig through tables and tables of numbers if you can see all the patterns in a single beautiful curve or a scatterplot? Variables relies heavily on quality graphics to aid you in the analysis, but of course there is always an option to check the exact values for your report or to confirm your intuition.

Data awareness

Data comes in many forms and kinds. And depending on its nature it can serve completely different purpose in the analysis. Variables knows that if the data represents counts of objects, rather than a continuous measure, it likely exhibits Poisson distribution and needs to be treated accordingly. Other variables, like text or dates, cannot be directly incorporated in the model but are employed to put other data in context.

Autosave

You never need to worry about saving your documents as Variables constantly keeps track of the changes you make and makes sure you never lose your work.

Import data

Variables can import CSV that can be exported from virtually any software. Copy from iTunes or open a file in Mail, Safari or any other app on your iPad.

Export plots and tables

Share the results of your work by exporting plots as vector PDF or raster PNG images. You can export tables too for pasting in a report or email.

Any orientation

You can use Variables in portrait or landscape, and it will adjust the layout to the way you like to hold your iPad.

Model creation

• Create models by directly manipulating variables and dependencies
• Awareness of model structure and data kinds to suggest optimal structure
• Selection of a subset of a model to check basic statistics
• Animated visualizations to see direct impact of changes made to a model
• Multiple models per dataset
• Highly optimized algorithms for matrix algebra and Fourier transforms

Available methods

• Descriptive statistics
• Histograms
• Kernel Density Estimates
• Correlations and covariances
• Scatterplots
• All of the above methods separated by categories of another variable
• Linear regression
• Logistic regression
• Poisson regression
• Model summaries and per-predictor statistics
• Actual vs Fitted plots
• Residuals vs Fitted plots

Visualisation

• Every chart or table can be enlarged and become interactive
• Great clean design of the charts with zero data obstruction
• Great color selection for multiple charts
• Zooming charts
• Panning charts
• Tapping a point in chart to learn more about the observation
• Export charts to PDF or PNG
• Export tables to RTF or plain text

Data management

• Import data from any CSV file
• Import from computer via iTunes or directly from Mail, Safari or any other app
• Automatic data kind recognition
• Renaming the variables, tables and models
• Adding variable descriptions and table attributions