修复数据清洗时间轴填补后的对齐问题

app/algorithms/api_ex/pressure_data_clean.py

@@ -37,8 +37,9 @@ def fill_time_gaps(
         start=data_indexed.index.min(), end=data_indexed.index.max(), freq=freq
     )
 
-    # 重索引以补齐缺失时间点
-    data_reindexed = data_indexed.reindex(full_range)
+    # 重索引以补齐缺失时间点，同时保留原始时间戳
+    combined_index = data_indexed.index.union(full_range).sort_values().unique()
+    data_reindexed = data_indexed.reindex(combined_index)
 
     # 按列处理缺口
     for col in data_reindexed.columns:
@@ -49,12 +50,12 @@ def fill_time_gaps(
         missing_groups = (is_missing != is_missing.shift()).cumsum()
         gap_lengths = is_missing.groupby(missing_groups).transform("sum")
 
-        # 短缺口：线性插值
+        # 短缺口：时间插值
         short_gap_mask = is_missing & (gap_lengths <= short_gap_threshold)
         if short_gap_mask.any():
             data_reindexed.loc[short_gap_mask, col] = (
                 data_reindexed[col]
-                .interpolate(method="linear", limit_area="inside")
+                .interpolate(method="time", limit_area="inside")
                 .loc[short_gap_mask]
             )
 
@@ -213,6 +214,12 @@ def clean_pressure_data_df_km(data: pd.DataFrame, show_plot: bool = False) -> di
     data_filled = fill_time_gaps(
         data, time_col="time", freq="1min", short_gap_threshold=10
     )
+
+    # 保存 time 列用于最后合并
+    time_col_series = None
+    if "time" in data_filled.columns:
+        time_col_series = data_filled["time"]
+
     # 移除 time 列用于后续清洗
     data_filled = data_filled.drop(columns=["time"])
 
@@ -304,6 +311,10 @@ def clean_pressure_data_df_km(data: pd.DataFrame, show_plot: bool = False) -> di
         plt.legend()
         plt.show()
 
+    # 将 time 列添加回结果
+    if time_col_series is not None:
+        data_repaired.insert(0, "time", time_col_series)
+
     # 返回清洗后的字典
     return data_repaired