If you need functionality that the built-in operators can't provide, you can create a user operator. With user operators, you can create new transformations, train models for future use, and make custom visualizations.
User operators can be viewed and edited in the Operators tab in the main menu. By default, this tab is hidden, but can be shown by toggling the Show operators menu option in your user profile settings.
Einblick's Own User Operators
Many of the operators you encounter in a canvas are actually implemented as user operators by Einblick's development team. These operators are kept in Einblick's user operator repository and are automatically updated in the platform whenever an update is made in the repository. Clicking the View source option on any of there operators in a canvas will take you to the corresponding page in the repository corresponding to the operator's underlying description and code. You can consult this repository and the user operator documentation for details on how to create and synchronize your own user operators.
User operators are a powerful way to extend the functionality of Einblick, and the following examples highlight some of the possibilities available to users that wish to create their own operators.
Data cleaning: Fill NaN values
Fill NaN values operator is a user operator, and removes any non-numeric values from the selected columns. In the image below, a dataframe is shown in a table before and after processing by the
Fill NaN values operator.
K-means clustering operator is also a user operator, and uses
MiniBatchKMeans to efficiently determine clusters in data based on the columns provided. Below, a dataframe is processed through the
K-means clustering operator and then visualized.
Linear regression operator runs a linear regression and returns both statistical information on the regression and an executor which is an operator that applies the linear regression prediction model on new data.
This operator creates a word cloud from a text column in the input dataframe.