Welcome to the crandas documentation!

Roseman Labs is software for secure data collaboration, based on multi-party computation (MPC). Crandas is a DataFrame library made to interact with the Roseman Labs cryptographic engine. With crandas, users can interact with data using MPC in a way familiar to data scientists without having to worry about the underlying cryptography.

Crandas can do the following:

• Data filtering and tabular manipulation

• Combination of different data sources

• Secure aggregations and other data analyses

Crandas also has an ever-growing library of privacy-preserving machine learning algorithms:

• Linear and logistic regression

• Random forests and k-nearest neighbors

• Hypothesis testing and other descriptive statistics

If you are new to crandas, the Getting Started guide will walk you through the process of installation up to basic data manipulation. If you are familiar with other DataFrame libraries, such as pandas, you might want to read this to find the essential differences.

For more in-depth information about functionalities, consult the User Guide. Additionally, you can go through some Tutorials to familiarize yourself with the different features present in crandas. Finally, for more generic information about Roseman Labs, the components it consists of and the role that crandas has as a part of this, visit the product page of the Roseman Labs Help Center.

Getting started

• Getting Started

User Guide

• Introduction
• Data Manipulation and Aggregation
• Machine Learning and Statistics
• Authorization

Reference

• API documentation
• Known limitations

Tutorials

• Tutorials

Index

• Index

• Index of functions per column type