12 Beyond R

Julia, Scala, and Java in the same notebook toolkit

Keywords

snowflake, R, RStudio, Posit, VS Code, workspace notebooks, snowflakeR, RSnowflake, mlops

12.1 Overview

Part IV is a short bridge: the same snowflake-notebook-multilang toolkit that powers %%R also supports Julia, Scala, and Java in Workspace Notebooks — shared micromamba, JDK, and EAI infrastructure.

You may never need these languages; this chapter exists so polyglot teams know the pattern exists and where to find docs.

Note

The blog series title “Beyond Python” describes the toolkit’s origin. In this guide, “Beyond R” means stepping outside R briefly before the RSnowflake and snowflakeR chapters.

12.2 Learning Objectives

Name the bridge library for each non-R language
Enable additional languages in YAML config
Point colleagues to sfnb-multilang docs vs this R-focused guide

12.3 Language matrix

All languages run on the Python kernel via cell magic — same model as %%R:

Language	Cell magic	Bridge	Typical use on Snowflake
R	`%%R`	rpy2	Stats, forecasting, snowflakeR ML — this guide
Scala	`%%scala`	JPype + Ammonite	Snowpark JVM, Spark Connect
Java	`%%java`	JPype	JVM libraries, Snowpark
Julia	`%%julia`	JuliaCall	Research prototypes, optimization

Each magic registers during setup_notebook() when the language is enabled in YAML.

12.4 Shared bootstrap

from sfnb_setup import setup_notebook

setup_notebook(
    config="multilang_config.yaml",
    packages=[],   # languages declared in YAML
)

Example YAML excerpt:

languages:
  r:
    enabled: true
  scala:
    enabled: true
    java_version: "17"
    snowpark_version: "1.18.0"
  julia:
    enabled: true
    julia_packages:
      - DataFrames
      - Plots

EAI must allow download hosts for each enabled language (JVM JARs, Julia packages, etc.). Bootstrap aggregates required domains.

Further reading: adding_a_language.md

12.5 Relationship to this guide

Resource	Focus
This guide	R + RSnowflake + snowflakeR + MLOps on Snowflake
snowflake-notebook-multilang	Equal-weight multilanguage bootstrap toolkit
Blog series `00_multilang_notebooks`	Narrative for Julia/Scala alongside R

snowflakeR and RSnowflake are R-only packages. Scala/Java/Julia cells do not use them — they use Snowpark JVM or native Julia ODBC patterns as documented in sfnb-multilang.

12.6 When to add another language

Scenario	Suggestion
Team already standardized on R for ML	Stay with `%%R` + snowflakeR
Spark/Snowpark JVM pipeline in same repo	Enable `%%scala` for shared notebook
Julia prototype alongside R production	Enable Julia in dev notebooks only
Greenfield Snowflake ML	Prefer R or Python consistently for maintainability

12.7 Next steps

The Snowflake SQL API — how RSnowflake connects.

--- title: "Beyond R" subtitle: "Julia, Scala, and Java in the same notebook toolkit" --- ## Overview Part IV is a short bridge: the same **snowflake-notebook-multilang** toolkit that powers `%%R` also supports **Julia**, **Scala**, and **Java** in **Workspace Notebooks** — shared micromamba, JDK, and EAI infrastructure. You may never need these languages; this chapter exists so polyglot teams know the pattern exists and where to find docs. ::: {.callout-note} The blog series title *"Beyond Python"* describes the toolkit's origin. In *this* guide, **"Beyond R"** means stepping outside R briefly before the RSnowflake and snowflakeR chapters. ::: ## Learning Objectives - Name the bridge library for each non-R language - Enable additional languages in YAML config - Point colleagues to sfnb-multilang docs vs this R-focused guide --- ## Language matrix {#sec-matrix} All languages run on the **Python kernel** via cell magic — same model as `%%R`: | Language | Cell magic | Bridge | Typical use on Snowflake | |----------|------------|--------|--------------------------| | **R** | `%%R` | rpy2 | Stats, forecasting, snowflakeR ML — **this guide** | | **Scala** | `%%scala` | JPype + Ammonite | Snowpark JVM, Spark Connect | | **Java** | `%%java` | JPype | JVM libraries, Snowpark | | **Julia** | `%%julia` | JuliaCall | Research prototypes, optimization | Each magic registers during `setup_notebook()` when the language is enabled in YAML. --- ## Shared bootstrap {#sec-shared} ```python from sfnb_setup import setup_notebook setup_notebook( config="multilang_config.yaml", packages=[], # languages declared in YAML ) ``` Example YAML excerpt: ```yaml languages: r: enabled: true scala: enabled: true java_version: "17" snowpark_version: "1.18.0" julia: enabled: true julia_packages: - DataFrames - Plots ``` EAI must allow download hosts for each enabled language (JVM JARs, Julia packages, etc.). Bootstrap aggregates required domains. **Further reading:** [adding_a_language.md](https://github.com/Snowflake-Labs/snowflake-notebook-multilang/blob/main/docs/adding_a_language.md) --- ## Relationship to this guide {#sec-relationship} | Resource | Focus | |----------|-------| | **This guide** | R + RSnowflake + snowflakeR + MLOps on Snowflake | | **snowflake-notebook-multilang** | Equal-weight multilanguage bootstrap toolkit | | **Blog series `00_multilang_notebooks`** | Narrative for Julia/Scala alongside R | snowflakeR and RSnowflake are **R-only** packages. Scala/Java/Julia cells do not use them — they use Snowpark JVM or native Julia ODBC patterns as documented in sfnb-multilang. --- ## When to add another language {#sec-when} | Scenario | Suggestion | |----------|------------| | Team already standardized on R for ML | Stay with `%%R` + snowflakeR | | Spark/Snowpark JVM pipeline in same repo | Enable `%%scala` for shared notebook | | Julia prototype alongside R production | Enable Julia in dev notebooks only | | Greenfield Snowflake ML | Prefer R or Python consistently for maintainability | --- ## Next steps [The Snowflake SQL API](../11_rsnowflake_sql_api/index.qmd) — how RSnowflake connects.