Key Terms¶
Here’s a list of key terms used in the Modelshop Platform.
Data¶
Import Wizard¶
The import wizard is how you import data into Modelshop. You can import data using a file, database, or web service. Once you import data, it becomes a datalist.
Data Connectors¶
Data connectors are your data sources. Data and calculations are separate, but together in Modelshop. You can refresh or reset your data without impacting your calculations. This also gives you the capability to add new data (say, you have new data from this quarter that you’d like to analyze) without having to rebuild your calculations.
Datalist¶
Datalists are tabular data structures that consists of rows (records) and columns (fields). Datalist fields can be raw data, calculated values, or links to other Datalist records. Additionally, Datalist fields can be arrays.
Data Types¶
Text: A string of characters not meant to represent a numerical value. Common examples are words, sentences, and phrases. The text data type can include alphabetic, numeric and special characters.
Integer: A whole number, including 0 and negative numbers. Integers cannot include fractions or decimals. Integers are often used to represent numbers that cannot have fractional values, like counts.
Floats and Decimals: A number that can have fractional values. Floats can also contain 0 and negative numbers.
Boolean: A value containing only true or false.
Date: A calendar date values. Hypothetically, dates could be stored as strings, however, the Date data type is preferable because it allows programmatic access to dates, including the day of the week, year, and to perform calculations between dates easily and efficiently. See more about handling dates, here.
Datetime: Similar with the Date Data Type, with the addition of a timestamp including hour, minute and second information.
Currency: Similar to a float, made to properly manage the specifics of currencies.
Datalist View¶
Views are persistent filters that can be applied to Datalists. Options include hiding certain fields, filtering records, grouping records, field formatting and masking. Additionally, permissions can be granted on a per view basis, for fine tuned security controls. Finally, each view becomes an independent API endpoint, which adheres to any options that have been applied for highly customized APIs.
Calculate¶
Field¶
Each column header in a datalist represents a field. Fields are where you place your calculations. Calculations can reference any other fields, from any datalists, and can use standard algebra or any of our preset expressions.
Field Editor¶
The field editor is where you add calculations. You can open the field editor by editing an existing field or by creating a new field. There are preset expressions in the field editor. You can choose from the dropdown menu for fields, methods, queries, functions, logic, and operators. You can also start typing and recommendations will pop up. Alternatively, you can use Java syntax in the field editor if that’s your preference.
Link¶
Links join data that have common fields. Similar to a “vlookup”, links help define relationships and query data that meets a specific criteria. Unlike a standard query though, you can link fields from all data sources and formats. Linking data automatically produces a new field that you can build calculations on top of. Using a link in a calculation gives you the power to reference an entire data set in a single cell. In each datalist, all linked data becomes a hyperlink. If you click on the hyperlink, you’re brought to the canvas where the two fields are highlighted. This helps you quickly identify which fields are connected. Unlike spreadsheets, there’s no limit on how many links or dimensions you can create in a model.
Link Wizard¶
The link wizard is where you join data together. The link wizard suggests links. You can also manually pick which datalists and fields you would like to link. The link wizard can be found in the top navigation or canvas.
Dependencies¶
Dependencies are how calculations relate to each other. They can be viewed on the canvas.
Pivot¶
Pivots are summary tables. Select fields to represent your facts (x axis) and dimensions (y axis). You can also filter your pivot by specific variables. Unlike spreadsheets, once you build a Pivot, each field on the Pivot table becomes a new variable that can be used in new calculations. It also empowers you to
Rule¶
A rule is a custom expression, that when triggered perform a defined action, such as prompting an alert or making a change to the model.
Rulebase¶
A Rulebase is a collection of rules.
Analytic¶
An analytic is how you train and optimize your model using an out-of-the-box machine learning algorithms. You can select “add analytic” in the navigation or canvas. You can also reference an analytic in the field editor.
Report¶
Chart¶
Charts are a visual representation of your data and calculations.
Dashboard¶
Dashboards are reports that you can use to help track and measure your business goals. You can create unlimited dashboards with custom charts. You can share dashboards with both internal and external collaborators, even those who do not have a Modelshop account.
Export¶
Export your model by downloading it into a zip file or exporting to a specific folder on your computer. You can chose to export the model in full, just the configuration, or the structure.
Manage your Model¶
Canvas¶
The canvas is a live graphical representation of your model. It is a drag and drop interface where you can import data, add calculations, create links, add predictive analytics, create charts, and publish your model. It’s a step-by-step guide of how to build a model.
Administration¶
Invite team members to create their own Modelshop accounts. You can set access to specific models and data sources depending on your business needs.
Modelshop APIs¶
The Modelshop API is a rest API that can be used for both importing data into the Modelshop platform and exporting your results, in real-time, to an external system.