Dataprep on GCP

09 March 2021

Cloud Dataprep is a Trifacta-operated integrated partner service focused on their industry-leading approach for data preparation. Google works closely with Trifacta to have a seamless user interface that eliminates the need for installation of up-front applications, separate licencing costs, or constant overhead operating costs. To meet your growing data preparation needs, Cloud Dataprep is completely managed and scales on demand so that you can remain focused on research.

• Fast exploration and anomaly detection
• Easy and powerful data preparation
• Predictive transformation
• Rich transformations
• Collaboration
• Active profiling
• Sampling
• Common data types
• Increased connectivity

Visual data distributions immediately understand and explore data. Cloud Dataprep identifies schemas, data forms, potential joins, and anomalies automatically, such as missing values, outliers, and duplicates, so that you can avoid the time-consuming task of determining the accuracy of the data and go to the discovery and review.

Cloud Dataprep automatically recommends and predicts the next ideal data transformation with each gesture in the UI. Cloud Dataprep uses Cloud Dataflow under the hood once you have established your sequence of transformations, allowing you to process structured or unstructured datasets of any size with the ease of clicks, not code.

It can be useful to be able to have several users work on the same assets in team environments or to make copies of good quality work to act as models for others.Cloud Dataprep allows users to collaborate in real-time on an equivalent flow objects or to make copies for individual work for others to use.

To help in exploration, cleansing, and transformation, see and explore your data through immersive visual distributions of your data. Visual representations help view vast quantities of data, and in a dynamic, easy-to-consume format, Cloud Dataprep’s creative profiling techniques visualise key statistical information.

Cloud Dataprep automatically produces one or more samples of the info for display and manipulation within the client application for performance optimization. You can however, easily adjust the sample size, the sample scope, and the process by which the sample is produced.

Schedule on a recurrent or as-needed basis the implementation of recipes in your flows. You will collect the wrangled output at the required output location when the scheduled job is successfully executed, where it is available in a published form that you specify.

Process data stored in Cloud Storage and with Cloud Identity and Access Control, user access and data protection are handled seamlessly.

In addition to the regular networking of BigQuery, Cloud Storage, Microsoft Excel and Google Sheets, enrich your self-service analytics with data sources from Salesforce, Oracle, Microsoft SQL Server, MySQL, and PostgreSQL.

In sequential and conditional order improve the automation capabilities by chaining data preparation jobs together. Warn users of failure or achievement and cause external tasks. Use comprehensive APIs as part of the end-to-end solution of an enterprise to incorporate Cloud Dataprep.

Expand on existing security requirements by using a blend of Google IAM positions and BigQuery, Cloud Storage, and Google Sheets access rights to establish access by providing user data access control.

On Google Cloud Console, navigate to https://console.cloud.google.com/dataprep and enable billing. The project owner is asked to enable data access by Google and Trifacta when you first open Dataprep. Then must approve the terms of service on that page, log in to their Google account and select the bucket from Cloud Storage and use it with Dataprep.

• Create a flow
• Import datasets
• Wrangle the Candidate file