Secure spaces for your statistical analysis and data science. R, RStudio, Python and Jupyter available in a compliant and isolated environment.
Access pre-configured analytical environments with all the packages needed for your research projects.
Complete statistical environment with pre-installed CRAN packages (survival, MatchIt, tableone, ggplot2, tidyverse and many others).
Data science, machine learning and analysis orchestration via DAGs. Pandas, scikit-learn, tensorflow, pytorch, matplotlib, seaborn, statsmodels. Jupyter also serves as the entry point for building and triggering Atlas workflows.
Docker container deployment orchestrated by Kubernetes (K-MANAGED cegedim.cloud). GitLab CI/CD pipeline with automatic security scanning before production release.
Atlas is the Python library developed by Clinityx to standardize, version and capitalize on statistical analysis code. Each package is a reusable, documented and tested micro-service.
Missing data management, multiple imputation, propensity scores (MICE).
Survival models (Kaplan-Meier, Cox, Fine & Gray), logistic, Poisson and log-binomial regressions.
Calculation of standardized incidence and prevalence.
Automatic generation of descriptive tables in Excel and HTML format.
Creation and assembly of inclusion/exclusion criteria to define study populations.
Feasibility
Assessment of available data
Planning
Selection of Atlas packages
Design
Analysis DAG in Jupyter
Implementation
execution and peer review
Merge
Integration into the study repository
Secure bubble completely isolated from the internet network. No external access possible from the analysis environments.
Environments completely disconnected from the internet. Impossible to download non-approved packages or communicate externally.
Data never leaves the analytical environment. Controlled, nominative and traced extractions. Complete audit trail of all accesses.
All actions are recorded (who has access, when, what, results). GDPR and HDS compliance guaranteed with 5-year logs.
Access via unique credentials. SSO support. MFA (multi-factor authentication) available for sensitive projects.
HDS certified
Infrastructure hosted in a Health Data Host (HDS) certified data center.
CNIL Authorized
Environments compliant with CNIL authorizations. Privacy by design implemented.
AES-256 encryption
Data encrypted at rest and in transit. Secure cryptographic keys.
Explore your health data with the best analysis tools in a completely secure and compliant environment.