neqsim_in_writing

NeqSim integration for scientific writing

This skill is loaded by book_author.agent and solve.task.agent whenever a chapter or paper section must defend a numerical statement with a NeqSim artifact. It complements book_creation (which handles structural authoring) and journal_formatting (which handles paper layout).

The objective is traceability: a reader (or reviewer) following any number, table, or figure in the manuscript must be able to land on a runnable NeqSim artifact in the repo within one click.

1. Standard chapter setup cell (Jupyter)

Every chapter notebook must start with the dual-boot cell. It tries the local devtools build first (so an editor sees fresh code) and falls back to the published neqsim PyPI package on Colab / external readers.

import importlib, subprocess, sys

try:
    from neqsim_dev_setup import neqsim_init, neqsim_classes
    ns = neqsim_init(recompile=False)
    ns = neqsim_classes(ns)
    NEQSIM_MODE = "devtools"
except Exception:
    try:
        import neqsim
    except ImportError:
        subprocess.check_call([sys.executable, "-m", "pip", "install", "-q", "neqsim"])
    from neqsim import jneqsim
    NEQSIM_MODE = "pip"

print(f"NeqSim loaded via {NEQSIM_MODE}")

Canonical imports inside the chapter body:

from neqsim import jneqsim
SystemSrkEos     = jneqsim.thermo.system.SystemSrkEos
SystemSrkCPAstatoil = jneqsim.thermo.system.SystemSrkCPAstatoil
ThermodynamicOperations = jneqsim.thermodynamicoperations.ThermodynamicOperations
ProcessSystem    = jneqsim.process.processmodel.ProcessSystem

Never instantiate JPype classes ad-hoc with JClass("...") inside book chapters — readers cannot easily spot the class. Use the named bindings above.

2. Claim → test linkage

Every assertion in prose that contains a number, equation, or qualitative behavioral statement that depends on simulation MUST carry a comment marker linking it to a JUnit test or a regression baseline.