Projects
There are four different project types in Arria Studio, each offering a different way of working with your data:
The first three take a table of data — for example, a table created in a spreadsheet program and saved in CSV format. The fourth type works with data in JSON format, which allows for a more varied and flexible data structure than a table.
Generally, your choice of project type will depend on how you want to use your data to construct a narrative. To make this choice easier, each project type is described below.
Describe Each Row
In this project type, each row in your input data is used to generate a separate story.
Suppose you have a table of employee data in which there is a single row for each company employee. Using this project type, you could generate a separate story for each employee. One limitation of this project type is that each story is self-contained — for example, the story generated for Row 4 of the table can refer to the data in Row 4 only.
For a step-by-step guide to building this type of project, see our Describing each row tutorial.
Describe the Table
In this project type, you generate a single story that corresponds to your whole data table. This project type is best when your data relates to a single entity with each row representing a constituent part of that entity.
Suppose you have a table of sales data with each row containing the sales data for a particular month. Using this project type, you could generate a story summarizing the sales data for the entire year. You could also focus on a specific region of the table — for example, you could focus on the data for January, February, and March to generate a story (or sentence within a story) about sales in Q1.
In this project type, the first column is treated as a unique identifier for the row; therefore, each value in the first column must be unique.
For a step-by-step guide to building this type of project, see our Describing a whole table tutorial.
Describe Row in Context
In this project type, each row in your input data is used to generate a story but you can add context to each story by referring to data in other rows. This allows you to create more sophisticated stories than is possible in Describe Each Row projects.
Suppose you have a table of crime data, with each row containing important crime statistics for a particular region of a country. Using this project type, you could generate a story for each region that (1) reports the crime statistics for that region, and (2) contextualizes the statistics by comparing them to the national average or the statistics for another region.
In this project type, the first column is treated as a unique identifier for the row; therefore, each value in the first column must be unique.
For a step-by-step guide to building this type of project, see our Describing rows in context tutorial.
Describe a JSON Object
In this project type, an entire JSON data structure is used to generate a single story. The data source can be any valid JSON object, allowing you to use a more complex data structure than a two-dimensional table.
To use this project type, you must know how JSON structures work. You must also learn about the special variable (WholeJSON) and the numerous data access functions that exist specifically for interrogating a JSON data structure.
For a step-by-step guide to building a project of this type, see our Describing a JSON object tutorial.
