July 30, 2023

Exploring the Latest Resource Management and Functional Improvements

We're excited to bring you some key updates to our Datrics platform. In this product update, we'll introduce our new deployment configurations - the Single Machine and Multiple Worker setups. But there's more! We've also made considerable enhancements to our Union Data brick and Pivot Table features. So, let's dive in and explore these updates, designed to optimize your resource management, streamline data preparation, and enhance data visualization.

Enhancements in Resource Management for Datrics On-Premises and SaaS Solutions

We are pleased to introduce two new deployment configurations for setting up Datrics in client environments - the Single Machine setup and the Multiple Worker setup. The selection between these two options can be tailored according to your organizational needs and requirements for optimal utilization.

The Multiple Worker setup comprises a single-node setup incorporating Docker containers. With this setup, various tasks like deployment runs, pipeline runs, and data previews are assigned to a dedicated worker. This arrangement enables more efficient resource utilization for data processing, model training, and deployment execution within your team.

Here are some practical applications of the Multiple Worker setup:

  • Pause workers during periods of inactivity, typically during night hours, to save on machine time costs.
  • For particularly resource-intensive tasks, such as model training, data scientists can initiate a larger worker and pause it when it's no longer required.
  • In larger teams, data scientists or analysts can create individual workers to distribute the load and prevent issues when executing multiple resource-intensive tasks simultaneously.
  • Allocate specific workers for deployments to ensure uninterrupted execution of tasks and prevent interference from experimental runs by data scientists.

The above-mentioned scenarios are just a few examples of how the Multiple Worker setup can be effectively utilized. These updates enhance the flexibility and efficiency of resource management in our platform.

Contact us to learn more.

Union multiple data frames

Sometimes to prepare the data one needs to merge data from multiple data sources or data outputs. Previously you could combine more than two data frames into one in multiple steps. In this release we are updating Union Data brick to concatenate any number of dataframes into one either vertically or horizontally.

Upgrades in Pivot Table Functionality

Furthermore, we have made significant improvements to our Pivot Table brick. This updated feature now retains its configuration across pipeline runs, providing greater convenience. You can establish the Pivot Table brick, execute the pipeline, and access the view without having to reset the configuration each time.

This brick is designed to build an interactive pivot table that summarizes and reorganizes selected columns and rows of data. This summary includes average, sum, count, standard deviation, and so on.

In addition, this brick supports multiple ways of visualization starting with regular tables and bar charts, ending with heatmap, pie, and line charts.

The full list of updates in the release.

Check out our previous updates

Datasets governance and Advanced settings for LGBM and XGBoost models | Datrics.ai

Datasets governance and Advanced settings for LGBM and XGBoost models

In this release we have enhanced the datasets and data sources management, and added monotone constraints to the LightGBM and XGBoost algorithms.
Datasets governance and Advanced settings for LGBM and XGBoost models | Datrics.ai

More opportunities for customization and predictive modeling

In this month’s updates we have focused on the modeling possibilities with no- and low-code. Firstly, neural prophet model was added to increase the time series forecasting capabilities in Datrics.
Datrics updates: New and updated data connectors in Datrics

New and updated data connectors in Datrics

In the latest product updates, we have redesigned the dataset creation screen, so that it’s more convenient to write queries and explore the uploaded data, added connectors to MongoDB and Redshift.
Do you want to discover more about Datrics?