Use existing loading state for sparse detail
This commit is contained in:
@@ -1,5 +1,7 @@
|
||||
# 概率训练样本归档说明(中文)
|
||||
|
||||
最后更新:`2026-04-18`
|
||||
|
||||
## 1. 目的
|
||||
|
||||
这份文档说明两件事:
|
||||
@@ -27,6 +29,12 @@
|
||||
- `max_so_far`
|
||||
- `peak_status`
|
||||
- `prob_snapshot`
|
||||
- `probability_engine`
|
||||
- `calibration_mode`
|
||||
- `calibration_version`
|
||||
- `raw_mu / raw_sigma`
|
||||
- `calibrated_mu / calibrated_sigma`
|
||||
- `shadow_distribution`
|
||||
- 当天最终 `actual_high`
|
||||
- 当天最终 `settlement bucket`
|
||||
|
||||
@@ -167,7 +175,10 @@
|
||||
],
|
||||
"probability_engine": "legacy",
|
||||
"probability_mode": "emos_shadow",
|
||||
"calibration_version": "emos-20260320130245"
|
||||
"calibration_mode": "emos_shadow",
|
||||
"calibration_version": "emos-20260320130245",
|
||||
"calibrated_mu": 15.4,
|
||||
"calibrated_sigma": 1.1
|
||||
}
|
||||
```
|
||||
|
||||
@@ -176,6 +187,8 @@
|
||||
- `actual_high`
|
||||
- `settlement_bucket`
|
||||
|
||||
当前前端把这类快照解释为“校准模型概率”。如果 `probability_engine` 为 LGBM 相关值,则显示为 LGBM 校准概率;模型舍入票数和市场价格只用于解释,不直接作为最终概率。
|
||||
|
||||
## 7. 现阶段你可以执行的命令
|
||||
|
||||
### 7.1 回填历史天气 CSV
|
||||
|
||||
Reference in New Issue
Block a user