Handle failures and retries

Debug from evidence. Reproduce one failure at a time, read the full exception or command output, and locate the first place actual state differs from expected state. Do not add retries to deterministic mistakes; fix the path, shape, name, or assumption.

Add .venv/ to .gitignore; environments contain machine-specific executables and consume unnecessary space. Review dependency updates and avoid installing untrusted packages with look-alike names.

python3 -m venv .venv
source .venv/bin/activate
python -m pip install pandas
python -m pip freeze > requirements.txt
python -c "import sys, pandas; print(sys.prefix); print(pandas.__version__)"

Expected evidence:

/project/.venv
2.x.x

The output may include version-specific details such as hashes, paths, fitted thresholds, or final decimal places. Compare the structural facts described here rather than copying placeholders. If the structure differs, stop and inspect the earliest unexpected line.

1. python3 -m venv .venv asks that exact interpreter to create the environment.
2. source changes shell variables so python resolves inside .venv on macOS and Linux.
3. python -m pip ties pip to the currently selected Python and avoids a mismatched executable.
4. freeze records resolved versions; the final command proves both interpreter location and import.

These lines form one chain: the interpreter version and declared package requirements becomes an isolated interpreter that imports the requested package at a recorded version. Change only one input first. When several values change together, you cannot tell which change caused the new behavior.

• First failure: an unchanged sys.prefix after activation means the activation command or shell path is wrong. Re-run the smallest command that proves the repair.
• Second failure: No module named pandas commonly means the editor or terminal is using a different interpreter. Preserve the failing input as a test when it represents a realistic mistake.
• Misleading success: a failing build after copying requirements.txt can indicate the file pins packages unavailable for the new Python or operating system. A clean-looking final line cannot cancel contradictory intermediate evidence.

When debugging, copy the exact error text and inspect names, paths, shapes, types, and versions. Explain the cause in one sentence before changing code. That discipline prevents a guessed repair from creating a second defect.

Keep the example small enough to inspect manually. Small does not mean careless: boundary values, file locations, feature order, and held-out data still determine whether the result means what you claim.

For the current debug step, save the smallest useful evidence: the relevant command, its output, and the input that produced it. Do not use a screenshot as the only record when text can be copied and searched. Keep generated artifacts separate from source inputs so rerunning the example does not destroy the evidence it is meant to evaluate.

The input contract is the interpreter version and declared package requirements. The visible result is an isolated interpreter that imports the requested package at a recorded version.

Return to your opening prediction. Mark it correct or rewrite it with the condition you missed. Then explain the difference between a successful execution and a trustworthy result for this specific example.