deduplicate power metrics

[rbx]

Environment already calculates proportional (core-based, not node-pased) power consumption metrics. No need to duplicate the work in the metrics code.

[coderabbitai]

📝 Walkthrough

Walkthrough

Removed proportional per-core power/cost computations and related metric fields across callbacks, metrics tracking, and reporting; replaced them with direct episode-level power metrics emitted from the environment and updated logging/printing to use non‑proportional totals and savings.

Changes

Cohort / File(s)	Summary
Callback & Logging src/callbacks.py	Dropped per-core proportional power/cost computation and imports of COST_IDLE_MW, COST_USED_MW, CORES_PER_NODE; log direct episode power metrics from env.metrics (metrics/power_mwh, metrics/bl_power_mwh, metrics/bl_off_power_mwh, metrics/savings_power_vs_baseline_off).
Metrics Computation src/metrics_tracker.py	Removed proportional-per-core power and cost calculations and all *_prop_* power/cost and proportional savings fields from record_episode_completion and returned episode_data; removed related config imports.
Reporting & Output src/evaluation_summary.py, train.py	Removed proportional-power fields from formatted summaries and console prints; replaced proportional power/cost reporting and savings computations with totals (*_power_mwh, *_cost) and percent-of-baseline computed from total savings vs baseline cost.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• Pr2 drop #44: Touches the same metrics pipeline and the same metric keys (*_prop_* vs totals) across callbacks, metrics_tracker, evaluation_summary, and train.
• Pr1 reward #40: Previously introduced proportional per-core power/cost computations and metric fields that this PR removes.

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'deduplicate power metrics' directly and concisely summarizes the main change: removing duplicate power metric calculations.
Description check	✅ Passed	The description clearly explains the rationale for the changes: the environment already calculates proportional power metrics, so duplication in the metrics code is unnecessary.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/evaluation_summary.py`:
- Around line 67-73: The episode summary currently emits Power and Savings twice
by repeating the same f-string block using episode_data (the lines building
f"Power={float(episode_data['agent_power_consumption_mwh'])... " and
f"Savings=€{float(episode_data['savings_vs_baseline'])..."); remove the
duplicate occurrence so the summary only formats and prints Power and Savings
once, keeping the original formatting and references to episode_data intact
(locate the duplicated f-string block in the episode summary function and delete
the redundant block).

In `@train.py`:
- Line 384: The print call at the static header in train.py uses an unnecessary
f-string (print(f"...")), which triggers Ruff F541; update the statement in the
same scope where that header is printed (the print of "=== COST SAVINGS (TOTAL
OVER EVALUATION) ===") by removing the leading f so it becomes a normal string
literal without any interpolation.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 3f2d6ad0-8de5-4e99-82fe-704bacd4ea17

📥 Commits

Reviewing files that changed from the base of the PR and between 30bbf23 and aeb59c3.

📒 Files selected for processing (4)

• src/callbacks.py
• src/evaluation_summary.py
• src/metrics_tracker.py
• train.py

💤 Files with no reviewable changes (1)

• src/metrics_tracker.py

[coderabbitai]

🧹 Nitpick comments (1)

src/callbacks.py (1)

1-1: Consider removing unused MAX_QUEUE_SIZE import.

MAX_QUEUE_SIZE is imported but not used—the only reference at line 41 is commented out.

Suggested fix

-from src.config import EPISODE_HOURS, MAX_QUEUE_SIZE, MAX_NODES
+from src.config import EPISODE_HOURS, MAX_NODES

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/callbacks.py` at line 1, The import list in src/callbacks.py includes
MAX_QUEUE_SIZE which is unused (the only reference is commented out); remove
MAX_QUEUE_SIZE from the from src.config import ... statement so the module only
imports the used symbols (EPISODE_HOURS, MAX_NODES) to eliminate the unused
import warning and keep imports minimal.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/callbacks.py`:
- Line 1: The import list in src/callbacks.py includes MAX_QUEUE_SIZE which is
unused (the only reference is commented out); remove MAX_QUEUE_SIZE from the
from src.config import ... statement so the module only imports the used symbols
(EPISODE_HOURS, MAX_NODES) to eliminate the unused import warning and keep
imports minimal.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: a43a8f26-8a16-43c6-8ef8-494a4ab5c76f

📥 Commits

Reviewing files that changed from the base of the PR and between aeb59c3 and 071ad21.

📒 Files selected for processing (4)

• src/callbacks.py
• src/evaluation_summary.py
• src/metrics_tracker.py
• train.py

💤 Files with no reviewable changes (1)

• src/metrics_tracker.py

✅ Files skipped from review due to trivial changes (1)

• train.py

🚧 Files skipped from review as they are similar to previous changes (1)

• src/evaluation_summary.py