Big Data Code Changes for Full Stack Simulation Engine 2026

Big Data Code Changes for the Full‑Stack Simulation Engine

Objectives and Key Tasks

• Fix and refactor code for the Full‑Stack Simulation Engine to support different simulation runs.

• Understand the end‑to‑end setup and architecture of the full‑stack simulation engine, including:

  • Source code

  • Configuration settings

  • Data inputs

  • PySpark pipelines

  • Grid and execution environments

• Run and debug PySpark code both locally and in the target (shared or distributed) environment.

• Fix and update YAML configuration files and the associated Python files, and clearly understand all inputs and parameters used in the YAML files.

• Configure PyCharm for effective local development and testing, especially local testing using small sample datasets.

• Modify and extend YAML parameters as required by new logic, experiments, or use cases.

• Fix variable definitions, naming issues, and scope inconsistencies — this is the primary responsibility of the role.

• Understand regression modeling concepts used in the simulation — this is to understand the larger picture, not to redesign models.

• Understand variable definitions, dependencies, and end‑to‑end data flow within the simulation model — this is required to ensure correctness and reproducibility.

Technical Setup

• Configure the local development environment in PyCharm.

• Run PySpark jobs for simulation workloads.

• Validate YAML‑driven configuration pipelines and ensure correct parameter propagation.

Code & Configuration

• Refactor full‑stack simulation engine code for clarity, correctness, and reusability.

• Debug Python and YAML integration issues.

• Correct variable definitions, parameter mappings, and configuration usage.

Modeling & Logic

• Understand regression modeling techniques used in the simulations.

• Analyze how variables impact model outputs across different runs.

• Ensure correctness, consistency, and repeatability across simulation executions.

Configs for Running Notebook on Grids – different venv, auth, types of auth, errors

Setting up git SSH on masters and local system – git clone, git push, and common errors like matching ticket name or just restart and soft reset

Making the code work locally then on master systems – local on 5k, recreating sample as needed

Changing code on Copilot for different filtering on big data notebooks running on masters which have access to HDFS

Writing notes in MD and also using Copilot

Handshaking CSVs for different repos – CSVs act as a bridge – also understanding CSV definition

Understanding errors that can happen in multithreading, hard to locate for PySpark local setup

Starting and running afresh from git clone, auth, and the notebook on master systems

Git clone locally, installing venv locally, and then running unit tests on local data

Manual setting in PyCharm for making local unit tests work

Creating samples that will work with different kinds of tests locally, like 5k local files

First make the current unit tests work – settings like no coverage, others in pytest.ini

Adding new variables for local tests