Flag.py

from typing import Dict, Union, List, Optional
from datetime import timedelta

import numpy as np
from obspy import UTCDateTime
from pydantic import BaseModel, Field, validator
from enum import Enum


class FlagCategory(str, Enum):
    ARTIFICIAL_DISTURBANCE = "ARTIFICIAL_DISTURBANCE"
    GAP = "GAP"
    EVENT = "EVENT"
    OTHER = "OTHER"


class Flag(BaseModel):
    """
        Base class for flagging features in magnetic timeseries data.

        Flag example:
        ```
        automatic_flag = Metadata(
            created_by='ex_algorithm',
            start_time=UTCDateTime('2023-01-01T03:05:10'),
            end_time=UTCDateTime('2023-01-01T03:05:11'),
            network='NT',
            station='BOU',
            channel='BEH',
            category=MetadataCategory.FLAG,
            comment="spike detected",
            priority=1,
            data_valid=False,
            metadata= ArtificialDisturbance{
                "description": "Spike in magnetic field strength",
                "field_work": false,
                "corrected": false,
                "flag_category": ARTIFICIAL_DISTURBANCE,
                "artificial_disturbance_type": ArtificialDisturbanceType.SPIKE,
                "source": "Lightning",
                "deviation": None,
        }
    )
        ```
    """

    description: str = Field(..., description="Description of the flag")
    field_work: bool = Field(..., description="Flag signaling field work")
    corrected: int = Field(..., description="Corrected ID for processing stage")
    flag_category: FlagCategory = "OTHER"


class ArtificialDisturbanceType(str, Enum):
    SPIKES = "SPIKES"
    OFFSET = "OFFSET"
    ARTIFICIAL_DISTURBANCES = "ARTIFICIAL_DISTURBANCES"


class ArtificialDisturbance(Flag):
    """
    This class is used to flag artificial disturbances.

    Artificial disturbances consist of the following types:

    SPIKES = Single data points that are outliers in the timeseries.
    OFFSET = A relatively constant shift or deviation in the baseline magnetic field.
    ARTIFICIAL_DISTURBANCES = A catch-all for a continuous period of unwanted variations, may include multiple spikes, offsets and/or gaps.

    Attributes
    ----------
    artificial_disturbance_type:ArtificialDisturbanceType
        The type of artificial disturbance(s).
    source: str
        Source of the disturbance if known or suspected.
    deviation: float
       Deviation of an offset in nt.
    spikes: np.ndarray
        NumPy array of timestamps as UTCDateTime. Can be a single spike or many spikes.

    """

    artificial_disturbance_type: ArtificialDisturbanceType
    deviation: Optional[float] = None
    source: Optional[str] = None
    spikes: Optional[np.ndarray] = None

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.flag_category = "ARTIFICIAL_DISTURBANCE"


class Gap(Flag):
    """
    This class is used to flag gaps in data.

    A gap is a period where data is missing or not recorded.

    Attributes
    ----------
    cause: str
        Cause of gap, e.g., network outage.
    handling: str
        How the gap is being handled, e.g., backfilled.
    """

    cause: str = None
    handling: str = None

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.flag_category = "GAP"


class EventType(str, Enum):
    GEOMAGNETIC_STORM = "GEOMAGNETIC_STORM"
    GEOMAGNETIC_SUBSTORM = "GEOMAGNETIC_SUBSTORM"
    EARTHQUAKE = "EARTHQUAKE"
    OTHER = "OTHER"


class Event(Flag):
    """
    This class is used to flag an event of interest such as a geomagnetic storm or earthquake.

    Attributes
    ----------
    event_type : EventType
        The type of event.
    scale : str
        Geomagnetic storm scale or Richter scale magnitude.
    index : int
        Planetary K-index, DST index or some other index.
    url : str
        A url related to the event. Could be NOAA SWPC, USGS Earthquakes page or another site.
    """

    event_type: EventType
    index: int = None
    scale: str = None
    url: str = None

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.flag_category = "EVENT"


# More example usage:
timestamps_array = np.array(
    [
        UTCDateTime("2023-11-16T12:00:0"),
        UTCDateTime("2023-11-16T12:01:10"),
        UTCDateTime("2023-11-16T12:02:30"),
    ]
)

spikes_data = {
    "starttime": "2023-11-16 12:00:00",
    "endtime": "2023-11-16 12:02:30",
    "description": "Spikes description",
    "field_work": False,
    "corrected": 32265,
    "disturbance_type": ArtificialDisturbanceType.SPIKES,
    "source": "processing",
    "spikes": timestamps_array,
}

offset_data = {
    "description": "Offset description",
    "field_work": False,
    "corrected": 47999,
    "disturbance_type": ArtificialDisturbanceType.OFFSET,
    "source": "Bin change",
    "deviation": 10.0,
}
geomagnetic_storm_data = {
    "description": "Geomagnetic storm",
    "field_work": False,
    "corrected": 36999,
    "event_type": EventType.GEOMAGNETIC_STORM,
    "scale": "G3",
    "index": 7,
    "url": "https://www.swpc.noaa.gov/products/planetary-k-index",
}

spike_instance = ArtificialDisturbance(**spikes_data)
offset_instance = ArtificialDisturbance(**offset_data)

print(spike_instance.model_dump())
print(offset_instance.model_dump())