Skip to content

🔬 lens

Lab Environment Notebook System

Visit lenslab.io → | GitHub →

Overview

lens provides flexible, reproducible computational environments for researchers including Jupyter notebooks, RStudio Server, and VSCode Server. Designed specifically for computational research workflows with multi-language support and collaborative features.

Key Features

Multiple Environments

  • Jupyter - Classic Notebook, JupyterLab, and JupyterHub
  • RStudio Server - Full-featured R development environment
  • VSCode Server - Browser-based VSCode for remote development
  • Terminal Access - SSH and web-based terminals

Language Support

  • Python - Multiple versions with scientific stack
  • R - R 4.3+ with Tidyverse and Bioconductor
  • Julia - Julia 1.9+ for high-performance computing
  • Multi-Language - Use Python, R, and Julia in same notebook

Collaboration

  • Multi-User - JupyterHub for team access
  • Shared Projects - Collaborative workspaces
  • Version Control - Git integration
  • Real-Time Sharing - Share live sessions

Reproducibility

  • Container-Based - Consistent environments via Docker
  • Environment Files - conda/pip/renv reproducible configs
  • Snapshots - Save environment state
  • Version Tracking - Track package versions

Use Cases

Computational Research

Develop computational models and analyze data with Jupyter notebooks and interactive R sessions.

Data Analysis

Perform statistical analysis with RStudio and visualize results interactively.

Teaching and Workshops

Provide students with consistent notebook environments for computational courses.

Collaborative Projects

Enable teams to work together in shared computational environments.

Architecture

graph TB
    User[Researcher] -->|Access| Hub[JupyterHub]
    Hub -->|Spawn| Jupyter[Jupyter Server]
    Hub -->|Spawn| RStudio[RStudio Server]
    Hub -->|Spawn| VSCode[VSCode Server]

    Jupyter -->|Access| EFS[Shared EFS Storage]
    RStudio -->|Access| EFS
    VSCode -->|Access| EFS

    EFS -->|Backup| S3[S3 Backups]

    style Hub fill:#ffe1f5
    style Jupyter fill:#e1f5ff
    style RStudio fill:#e1ffe1
    style VSCode fill:#f5e1ff

Available Environments

Python Environments

  • Data Science - pandas, NumPy, scikit-learn, Matplotlib
  • Geospatial - GeoPandas, Rasterio, Shapely
  • Bioinformatics - Biopython, scanpy, AnnData
  • Machine Learning - PyTorch, TensorFlow (CPU)

R Environments

  • Tidyverse - dplyr, ggplot2, tidyr, readr
  • Bioconductor - DESeq2, edgeR, Seurat
  • Spatial - sf, terra, stars
  • Statistics - lme4, survival, caret

Julia Environments

  • Scientific Computing - DifferentialEquations.jl
  • Data Analysis - DataFrames.jl, Plots.jl
  • Optimization - JuMP.jl, Optim.jl

Getting Started

Launch an Environment

# Install lens CLI
brew install lens

# List available environments
lens list-environments

# Launch Jupyter environment
lens launch jupyter \
  --name my-notebook \
  --environment datascience-python

# Launch RStudio environment
lens launch rstudio \
  --name my-rstudio \
  --environment tidyverse

# Launch VSCode environment
lens launch vscode \
  --name my-vscode \
  --environment julia

Access Your Environment

# Get connection URL
lens url my-notebook

# Open in browser
lens open my-notebook

# Connect via SSH for tunneling
lens ssh my-notebook

Manage Environments

# List running environments
lens list

# Stop environment (preserves data)
lens stop my-notebook

# Start stopped environment
lens start my-notebook

# Export environment config
lens export my-notebook > environment.yml

# Create from config
lens create --from environment.yml --name reproduced

Pre-Installed Packages

Python Scientific Stack

# Core libraries
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Machine learning
import sklearn
import scipy

# Geospatial
import geopandas as gpd
import rasterio

R Tidyverse

# Core tidyverse
library(tidyverse)
library(readr)
library(dplyr)
library(ggplot2)

# Statistics
library(lme4)
library(survival)

# Bioconductor
library(DESeq2)
library(edgeR)

Julia

# Data analysis
using DataFrames
using CSV
using Plots

# Scientific computing
using DifferentialEquations
using Optim

Integration with ResearchComputing

Development Workflow

  1. Develop and test in lens notebooks
  2. Export code for batch execution
  3. Scale to atom for HPC runs
  4. Archive notebooks and results in cargoship

Data Pipeline

  • Load data from S3
  • Process in interactive notebooks
  • Export results back to S3
  • Retrieve archived data from cargoship

Account Provisioning

  • Accounts managed via petri
  • Cost tracking and budgets
  • Multi-user access control

Features

Jupyter Features

  • Extensions - JupyterLab extensions for productivity
  • Kernels - Python, R, Julia, and more
  • Widgets - Interactive widgets with ipywidgets
  • Export - Export to HTML, PDF, slides

RStudio Features

  • Projects - RStudio projects for organization
  • Git Integration - Built-in version control
  • Package Management - renv for reproducibility
  • Shiny - Deploy Shiny apps for interactive dashboards

VSCode Features

  • Extensions - Python, R, Julia extensions
  • Remote Development - Full VSCode experience
  • Debugging - Interactive debugging
  • Terminal - Integrated terminal

Documentation

Technology Stack

  • Environments: JupyterHub, RStudio Server, code-server
  • Containers: Docker, custom images
  • Storage: EFS (shared), EBS (local)
  • Authentication: JupyterHub authenticators, OAuth
  • Infrastructure: ECS, ALB, CloudFormation

Project Status

Current Version: v0.5.0 Status: Active Development License: Apache 2.0

Contributing

Contributions welcome! See the contribution guide.

Support

← Back to Ecosystem