REF: refactor ids_finder and process_events functions to improve para…

…meter handling

notebooks/00_ids_finder.ipynb

@@ -41,7 +41,7 @@
     "import polars as pl\n",
     "from discontinuitypy.detection.variance import detect_variance\n",
     "from discontinuitypy.core.propeties import process_events\n",
-    "from discontinuitypy.utils.basic import df2ts\n",
+    "from space_analysis.ds.ts.io import df2ts\n",
     "from loguru import logger\n",
     "\n",
     "from datetime import timedelta\n",
@@ -89,10 +89,9 @@
     "# | export\n",
     "def ids_finder(\n",
     "    detection_df: pl.LazyFrame,  # data used for anomaly dectection (typically low cadence data)\n",
-    "    tau: timedelta,\n",
-    "    ts: timedelta,\n",
     "    bcols=None,\n",
     "    detect_func: Callable[..., pl.LazyFrame] = detect_variance,\n",
+    "    detect_kwargs: dict = {},\n",
     "    extract_df: pl.LazyFrame = None,  # data used for feature extraction (typically high cadence data),\n",
     "    **kwargs,\n",
     "):\n",
@@ -108,11 +107,11 @@
     "    detection_df = detection_df.sort(\"time\")\n",
     "    extract_df = extract_df.sort(\"time\")\n",
     "\n",
-    "    events = detect_func(detection_df, tau, ts, bcols, **kwargs)\n",
+    "    events = detect_func(detection_df, bcols=bcols, **detect_kwargs)\n",
     "\n",
     "    data_c = compress_data_by_events(extract_df.collect(), events)\n",
     "    sat_fgm = df2ts(data_c, bcols)\n",
-    "    ids = process_events(events, sat_fgm, ts, **kwargs)\n",
+    "    ids = process_events(events, sat_fgm, **kwargs)\n",
     "    return ids"
    ]
   },

notebooks/02_ids_properties.ipynb

@@ -34,7 +34,6 @@
     "except ImportError:\n",
     "    import pandas as pd\n",
     "\n",
-    "from datetime import timedelta\n",
     "from loguru import logger\n",
     "\n",
     "from beforerr.polars import convert_to_pd_dataframe, decompose_vector\n",
@@ -388,7 +387,6 @@
     "def process_events(\n",
     "    candidates_pl: pl.DataFrame,  # potential candidates DataFrame\n",
     "    sat_fgm: xr.DataArray,  # satellite FGM data\n",
-    "    data_resolution: timedelta,  # time resolution of the data\n",
     "    method: Literal[\"fit\", \"derivative\"] = \"fit\",\n",
     "    **kwargs,\n",
     ") -> pl.DataFrame:\n",

src/discontinuitypy/core/pipeline.py

@@ -8,7 +8,7 @@
 import polars as pl
 from ..detection.variance import detect_variance
 from .propeties import process_events
-from ..utils.basic import df2ts
+from space_analysis.ds.ts.io import df2ts
 from loguru import logger
 
 from datetime import timedelta
@@ -28,10 +28,9 @@ def compress_data_by_events(data: pl.DataFrame, events: pl.DataFrame):
 # %% ../../../notebooks/00_ids_finder.ipynb 6
 def ids_finder(
     detection_df: pl.LazyFrame,  # data used for anomaly dectection (typically low cadence data)
-    tau: timedelta,
-    ts: timedelta,
     bcols=None,
     detect_func: Callable[..., pl.LazyFrame] = detect_variance,
+    detect_kwargs: dict = {},
     extract_df: pl.LazyFrame = None,  # data used for feature extraction (typically high cadence data),
     **kwargs,
 ):
@@ -47,11 +46,11 @@ def ids_finder(
     detection_df = detection_df.sort("time")
     extract_df = extract_df.sort("time")
 
-    events = detect_func(detection_df, tau, ts, bcols, **kwargs)
+    events = detect_func(detection_df, bcols=bcols, **detect_kwargs)
 
     data_c = compress_data_by_events(extract_df.collect(), events)
     sat_fgm = df2ts(data_c, bcols)
-    ids = process_events(events, sat_fgm, ts, **kwargs)
+    ids = process_events(events, sat_fgm, **kwargs)
     return ids
 
 # %% ../../../notebooks/00_ids_finder.ipynb 8

src/discontinuitypy/core/propeties.py

@@ -22,7 +22,6 @@
 except ImportError:
     import pandas as pd
 
-from datetime import timedelta
 from loguru import logger
 
 from beforerr.polars import convert_to_pd_dataframe, decompose_vector
@@ -254,7 +253,6 @@ def calc_normal_direction(data, name="normal_direction", **kwargs):
 def process_events(
     candidates_pl: pl.DataFrame,  # potential candidates DataFrame
     sat_fgm: xr.DataArray,  # satellite FGM data
-    data_resolution: timedelta,  # time resolution of the data
     method: Literal["fit", "derivative"] = "fit",
     **kwargs,
 ) -> pl.DataFrame: