feat: R-compatible summary format with full CausalImpactResults

[YuminosukeSato]

Summary

• Add 14 new fields to CausalImpactResults to match R CausalImpact summary output
• Replace "-" placeholders in Actual / Prediction (s.d.) rows with computed values
• Clamp all s.d. fields to 0 when n_samples=1 to prevent ddof=1 NaN

Changed files

File	Change
python/causal_impact/analysis.py	Add 14 fields to CausalImpactResults + cross-sample aggregation in compute_effects
python/causal_impact/summary.py	R-compatible format (Actual row, Prediction(s.d.) row, 3 CI rows, always show Posterior prob.)
tests/test_summary.py	Fix Absolute effect CI row index (7→8) due to new format layout

summary() output (Before → After)

Before:

Actual                   -              -
Prediction (s.d.)        -              -
95% CI                   [9.08, 10.98]  [272.39, 329.38]

After:

Actual                   15.02          450.67
Prediction (s.d.)        10.03 (0.48)   300.89 (14.40)
95% CI                   [9.08, 10.98]  [272.39, 329.38]

Absolute effect (s.d.)   4.99 (0.48)    149.78 (14.40)
95% CI                   [4.04, 5.94]   [121.29, 178.28]

Relative effect (s.d.)   49.8% (4.8%)   49.8% (4.8%)
95% CI                   [40.3%, 59.3%] [40.3%, 59.3%]

Posterior tail-area probability p: 0.001
Posterior prob. of a causal effect: 99.90%

Test plan

• test_analysis.py: 28 passed (2 previously Red tests now Green)
• test_summary.py: 8 passed (all 8 previously Red tests now Green)
• Full test suite: 180 passed
• ruff check / ruff format: passed