imap-processing 1.0.1__py3-none-any.whl → 1.0.3__py3-none-any.whl

imap_processing/codice/codice_l1a_lo_angular.py

@@ -0,0 +1,362 @@
+"""CoDICE Lo Angular L1A processing functions."""
+
+import logging
+from pathlib import Path
+
+import numpy as np
+import xarray as xr
+
+from imap_processing.cdf.imap_cdf_manager import ImapCdfAttributes
+from imap_processing.codice import constants
+from imap_processing.codice.decompress import decompress
+from imap_processing.codice.utils import (
+    CODICEAPID,
+    ViewTabInfo,
+    calculate_acq_time_per_step,
+    get_codice_epoch_time,
+    get_collapse_pattern_shape,
+    get_view_tab_info,
+    index_to_position,
+    read_sci_lut,
+)
+
+logger = logging.getLogger(__name__)
+
+
+def _despin_species_data(
+    species_data: np.ndarray, sci_lut_data: dict, view_tab_obj: ViewTabInfo
+) -> np.ndarray:
+    """
+    Apply despinning mapping for angular products.
+
+    Despinned data shape is (num_packets, num_species, 24, inst_az) where
+    we expand spin_sector to 24 by filling with zeros in 12 to 24 or 0 to 11
+    based on pixel orientation.
+
+    Parameters
+    ----------
+    species_data : np.ndarray
+        The species data array to be despun.
+    sci_lut_data : dict
+        The science LUT data used for despinning.
+    view_tab_obj : ViewTabInfo
+        The view table information object.
+
+    Returns
+    -------
+    np.ndarray
+        The despun species data array in
+        (num_packets, num_species, esa_steps, 24, inst_az).
+    """
+    # species_data shape: (num_packets, num_species, esa_steps, *collapsed_dims)
+    num_packets, num_species, esa_steps = species_data.shape[:3]
+    collapsed_dims = species_data.shape[3:]
+    inst_az_dim = collapsed_dims[-1]
+
+    # Prepare despinning output: (num_packets, num_species, esa_steps, 24, inst_az_dim)
+    # 24 is derived by multiplying spin sector dim from collapse table by 2
+    spin_sector_len = constants.LO_DESPIN_SPIN_SECTORS
+    despun_shape = (num_packets, num_species, esa_steps, spin_sector_len, inst_az_dim)
+    despun_data = np.full(despun_shape, 0)
+
+    # Pixel orientation array and mapping positions
+    pixel_orientation = np.array(
+        sci_lut_data["lo_stepping_tab"]["pixel_orientation"]["data"]
+    )
+    # index_to_position gets the position from collapse table. Eg.
+    #   [1, 2, 3, 23, 24] for SW angular
+    angular_position = index_to_position(sci_lut_data, 0, view_tab_obj.collapse_table)
+    orientation_a = pixel_orientation == "A"
+    orientation_b = pixel_orientation == "B"
+
+    # Despin data based on orientation and angular position
+    for pos_idx, position in enumerate(angular_position):
+        if position <= 12:
+            # Case 1: position 0-12, orientation A, append to first half
+            despun_data[:, :, orientation_a, :12, pos_idx] = species_data[
+                :, :, orientation_a, :, pos_idx
+            ]
+            # Case 2: position 12-24, orientation B, append to second half
+            despun_data[:, :, orientation_b, 12:, pos_idx] = species_data[
+                :, :, orientation_b, :, pos_idx
+            ]
+        else:
+            # Case 3: position 12-24, orientation A, append to second half
+            despun_data[:, :, orientation_a, 12:, pos_idx] = species_data[
+                :, :, orientation_a, :, pos_idx
+            ]
+            # Case 4: position 0-12, orientation B, append to first half
+            despun_data[:, :, orientation_b, :12, pos_idx] = species_data[
+                :, :, orientation_b, :, pos_idx
+            ]
+
+    return despun_data
+
+
+def l1a_lo_angular(unpacked_dataset: xr.Dataset, lut_file: Path) -> xr.Dataset:
+    """
+    L1A processing code.
+
+    Parameters
+    ----------
+    unpacked_dataset : xarray.Dataset
+        The decompressed and unpacked data from the packet file.
+    lut_file : pathlib.Path
+        Path to the LUT (Lookup Table) file used for processing.
+
+    Returns
+    -------
+    xarray.Dataset
+        The processed L1A dataset for the given species product.
+    """
+    # Get these values from unpacked data. These are used to
+    # lookup in LUT table.
+    table_id = unpacked_dataset["table_id"].values[0]
+    view_id = unpacked_dataset["view_id"].values[0]
+    apid = unpacked_dataset["pkt_apid"].values[0]
+    plan_id = unpacked_dataset["plan_id"].values[0]
+    plan_step = unpacked_dataset["plan_step"].values[0]
+
+    logger.info(
+        f"Processing angular with - APID: 0x{apid:X}, View ID: {view_id}, "
+        f"Table ID: {table_id}, Plan ID: {plan_id}, Plan Step: {plan_step}"
+    )
+
+    # ========== Get LUT Data ===========
+    # Read information from LUT
+    sci_lut_data = read_sci_lut(lut_file, table_id)
+
+    view_tab_info = get_view_tab_info(sci_lut_data, view_id, apid)
+    view_tab_obj = ViewTabInfo(
+        apid=apid,
+        view_id=view_id,
+        sensor=view_tab_info["sensor"],
+        three_d_collapsed=view_tab_info["3d_collapse"],
+        collapse_table=view_tab_info["collapse_table"],
+    )
+
+    if view_tab_obj.sensor != 0:
+        raise ValueError("Unsupported sensor ID for Lo angular processing.")
+
+    # ========= Decompress and Reshape Data ===========
+    # Lookup SW or NSW species based on APID
+    if view_tab_obj.apid == CODICEAPID.COD_LO_SW_ANGULAR_COUNTS:
+        species_names = sci_lut_data["data_product_lo_tab"]["0"]["angular"]["sw"][
+            "species_names"
+        ]
+        logical_source_id = "imap_codice_l1a_lo-sw-angular"
+    elif view_tab_obj.apid == CODICEAPID.COD_LO_NSW_ANGULAR_COUNTS:
+        species_names = sci_lut_data["data_product_lo_tab"]["0"]["angular"]["nsw"][
+            "species_names"
+        ]
+        logical_source_id = "imap_codice_l1a_lo-nsw-angular"
+    else:
+        raise ValueError(f"Unknown apid {view_tab_obj.apid} in Lo species processing.")
+
+    compression_algorithm = constants.LO_COMPRESSION_ID_LOOKUP[view_tab_obj.view_id]
+    # Decompress data using byte count information from decommed data
+    binary_data_list = unpacked_dataset["data"].values
+    byte_count_list = unpacked_dataset["byte_count"].values
+
+    # The decompressed data in the shape of (epoch, n). Then reshape later.
+    decompressed_data = [
+        decompress(
+            packet_data[:byte_count],
+            compression_algorithm,
+        )
+        for (packet_data, byte_count) in zip(
+            binary_data_list, byte_count_list, strict=False
+        )
+    ]
+
+    # Look up collapse pattern using LUT table. This should return collapsed shape.
+    collapsed_shape = get_collapse_pattern_shape(
+        sci_lut_data, view_tab_obj.sensor, view_tab_obj.collapse_table
+    )
+
+    # Reshape decompressed data to:
+    #   (num_packets, num_species, esa_steps, 12, 5)
+    # 24 includes despinning spin sector. Then at later steps,
+    # we handle despinning.
+    num_packets = len(binary_data_list)
+    esa_steps = constants.NUM_ESA_STEPS
+    num_species = len(species_names)
+    species_data = np.array(decompressed_data).reshape(
+        num_packets, num_species, esa_steps, *collapsed_shape
+    )
+
+    # Despinning
+    # ----------------
+    species_data = _despin_species_data(species_data, sci_lut_data, view_tab_obj)
+
+    # ========== Get Voltage Data from LUT ===========
+    # Use plan id and plan step to get voltage data's table_number in ESA sweep table.
+    # Voltage data is (128,)
+    esa_table_number = sci_lut_data["plan_tab"][f"({plan_id}, {plan_step})"][
+        "lo_stepping"
+    ]
+    voltage_data = sci_lut_data["esa_sweep_tab"][f"{esa_table_number}"]
+
+    # ========= Get Epoch Time Data ===========
+    # Epoch center time and delta
+    epoch_center, deltas = get_codice_epoch_time(
+        unpacked_dataset["acq_start_seconds"].values,
+        unpacked_dataset["acq_start_subseconds"].values,
+        unpacked_dataset["spin_period"].values,
+        view_tab_obj,
+    )
+
+    # ========== Create CDF Dataset with Metadata ===========
+    cdf_attrs = ImapCdfAttributes()
+    cdf_attrs.add_instrument_global_attrs("codice")
+    cdf_attrs.add_instrument_variable_attrs("codice", "l1a")
+
+    l1a_dataset = xr.Dataset(
+        coords={
+            "epoch": xr.DataArray(
+                epoch_center,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes("epoch", check_schema=False),
+            ),
+            "epoch_delta_minus": xr.DataArray(
+                deltas,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "epoch_delta_minus", check_schema=False
+                ),
+            ),
+            "epoch_delta_plus": xr.DataArray(
+                deltas,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "epoch_delta_plus", check_schema=False
+                ),
+            ),
+            "esa_step": xr.DataArray(
+                np.arange(128),
+                dims=("esa_step",),
+                attrs=cdf_attrs.get_variable_attributes("esa_step", check_schema=False),
+            ),
+            "esa_step_label": xr.DataArray(
+                np.arange(128).astype(str),
+                dims=("esa_step",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "esa_step_label", check_schema=False
+                ),
+            ),
+            "inst_az": xr.DataArray(
+                index_to_position(sci_lut_data, 0, view_tab_obj.collapse_table),
+                dims=("inst_az",),
+                attrs=cdf_attrs.get_variable_attributes("inst_az", check_schema=False),
+            ),
+            "inst_az_label": xr.DataArray(
+                index_to_position(sci_lut_data, 0, view_tab_obj.collapse_table).astype(
+                    str
+                ),
+                dims=("inst_az",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "inst_az_label", check_schema=False
+                ),
+            ),
+            "k_factor": xr.DataArray(
+                np.array([constants.K_FACTOR]),
+                dims=("k_factor",),
+                attrs=cdf_attrs.get_variable_attributes("k_factor", check_schema=False),
+            ),
+            "spin_sector": xr.DataArray(
+                np.arange(24, dtype=np.uint8),
+                dims=("spin_sector",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "spin_sector", check_schema=False
+                ),
+            ),
+            "spin_sector_label": xr.DataArray(
+                np.arange(24).astype(str),
+                dims=("spin_sector",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "spin_sector_label", check_schema=False
+                ),
+            ),
+        },
+        attrs=cdf_attrs.get_global_attributes(logical_source_id),
+    )
+    # Add first few unique variables
+    l1a_dataset["k_factor"] = xr.DataArray(
+        np.array([constants.K_FACTOR]),
+        dims=("k_factor",),
+        attrs=cdf_attrs.get_variable_attributes("k_factor_attrs", check_schema=False),
+    )
+    l1a_dataset["spin_period"] = xr.DataArray(
+        unpacked_dataset["spin_period"].values * constants.SPIN_PERIOD_CONVERSION,
+        dims=("epoch",),
+        attrs=cdf_attrs.get_variable_attributes("spin_period"),
+    )
+    l1a_dataset["voltage_table"] = xr.DataArray(
+        np.array(voltage_data),
+        dims=("esa_step",),
+        attrs=cdf_attrs.get_variable_attributes("voltage_table", check_schema=False),
+    )
+    l1a_dataset["data_quality"] = xr.DataArray(
+        unpacked_dataset["suspect"].values,
+        dims=("epoch",),
+        attrs=cdf_attrs.get_variable_attributes("data_quality"),
+    )
+    l1a_dataset["acquisition_time_per_step"] = xr.DataArray(
+        calculate_acq_time_per_step(sci_lut_data["lo_stepping_tab"]),
+        dims=("esa_step",),
+        attrs=cdf_attrs.get_variable_attributes(
+            "acquisition_time_per_step", check_schema=False
+        ),
+    )
+
+    # Carry over these variables from unpacked data to l1a_dataset
+    l1a_carryover_vars = [
+        "sw_bias_gain_mode",
+        "st_bias_gain_mode",
+        "rgfo_half_spin",
+        "nso_half_spin",
+    ]
+    # Loop through them since we need to set their attrs too
+    for var in l1a_carryover_vars:
+        l1a_dataset[var] = xr.DataArray(
+            unpacked_dataset[var].values,
+            dims=("epoch",),
+            attrs=cdf_attrs.get_variable_attributes(var),
+        )
+
+    # Finally, add species data variables and their uncertainties
+    for species_data_idx, species in enumerate(species_names):
+        species_attrs = cdf_attrs.get_variable_attributes("lo-angular-attrs")
+        unc_attrs = cdf_attrs.get_variable_attributes("lo-angular-unc-attrs")
+
+        direction = (
+            "Sunward"
+            if view_tab_obj.apid == CODICEAPID.COD_LO_SW_ANGULAR_COUNTS
+            else "Non-Sunward"
+        )
+        # Replace {species} and {direction} in attrs
+        species_attrs["CATDESC"] = species_attrs["CATDESC"].format(
+            species=species, direction=direction
+        )
+        species_attrs["FIELDNAM"] = species_attrs["FIELDNAM"].format(
+            species=species, direction=direction
+        )
+        l1a_dataset[species] = xr.DataArray(
+            species_data[:, species_data_idx, :, :, :],
+            dims=("epoch", "esa_step", "spin_sector", "inst_az"),
+            attrs=species_attrs,
+        )
+        # Uncertainty data
+        unc_attrs["CATDESC"] = unc_attrs["CATDESC"].format(
+            species=species, direction=direction
+        )
+        unc_attrs["FIELDNAM"] = unc_attrs["FIELDNAM"].format(
+            species=species, direction=direction
+        )
+        l1a_dataset[f"unc_{species}"] = xr.DataArray(
+            np.sqrt(species_data[:, species_data_idx, :, :, :]),
+            dims=("epoch", "esa_step", "spin_sector", "inst_az"),
+            attrs=unc_attrs,
+        )
+
+    return l1a_dataset

imap_processing/codice/codice_l1a_lo_species.py

@@ -0,0 +1,282 @@
+"""CoDICE Lo Species L1A processing functions."""
+
+import logging
+from pathlib import Path
+
+import numpy as np
+import xarray as xr
+
+from imap_processing.cdf.imap_cdf_manager import ImapCdfAttributes
+from imap_processing.codice import constants
+from imap_processing.codice.decompress import decompress
+from imap_processing.codice.utils import (
+    CODICEAPID,
+    ViewTabInfo,
+    calculate_acq_time_per_step,
+    get_codice_epoch_time,
+    get_collapse_pattern_shape,
+    get_view_tab_info,
+    read_sci_lut,
+)
+
+logger = logging.getLogger(__name__)
+
+
+def l1a_lo_species(unpacked_dataset: xr.Dataset, lut_file: Path) -> xr.Dataset:
+    """
+    L1A processing code.
+
+    Parameters
+    ----------
+    unpacked_dataset : xarray.Dataset
+        The decompressed and unpacked data from the packet file.
+    lut_file : pathlib.Path
+        Path to the LUT (Lookup Table) file used for processing.
+
+    Returns
+    -------
+    xarray.Dataset
+        The processed L1A dataset for the given species product.
+    """
+    # Get these values from unpacked data. These are used to
+    # lookup in LUT table.
+    table_id = unpacked_dataset["table_id"].values[0]
+    view_id = unpacked_dataset["view_id"].values[0]
+    apid = unpacked_dataset["pkt_apid"].values[0]
+    plan_id = unpacked_dataset["plan_id"].values[0]
+    plan_step = unpacked_dataset["plan_step"].values[0]
+
+    logger.info(
+        f"Processing species with - APID: {apid}, View ID: {view_id}, "
+        f"Table ID: {table_id}, Plan ID: {plan_id}, Plan Step: {plan_step}"
+    )
+
+    # ========== Get LUT Data ===========
+    # Read information from LUT
+    sci_lut_data = read_sci_lut(lut_file, table_id)
+
+    view_tab_info = get_view_tab_info(sci_lut_data, view_id, apid)
+    view_tab_obj = ViewTabInfo(
+        apid=apid,
+        view_id=view_id,
+        sensor=view_tab_info["sensor"],
+        three_d_collapsed=view_tab_info["3d_collapse"],
+        collapse_table=view_tab_info["collapse_table"],
+    )
+
+    if view_tab_obj.sensor != 0:
+        raise ValueError("Unsupported sensor ID for Lo species processing.")
+
+    # ========= Decompress and Reshape Data ===========
+    # Lookup SW or NSW species based on APID
+    if view_tab_obj.apid == CODICEAPID.COD_LO_SW_SPECIES_COUNTS:
+        species_names = sci_lut_data["data_product_lo_tab"]["0"]["species"]["sw"][
+            "species_names"
+        ]
+        logical_source_id = "imap_codice_l1a_lo-sw-species"
+    elif view_tab_obj.apid == CODICEAPID.COD_LO_NSW_SPECIES_COUNTS:
+        species_names = sci_lut_data["data_product_lo_tab"]["0"]["species"]["nsw"][
+            "species_names"
+        ]
+        logical_source_id = "imap_codice_l1a_lo-nsw-species"
+    else:
+        raise ValueError(f"Unknown apid {view_tab_obj.apid} in Lo species processing.")
+
+    compression_algorithm = constants.LO_COMPRESSION_ID_LOOKUP[view_tab_obj.view_id]
+    # Decompress data using byte count information from decommed data
+    binary_data_list = unpacked_dataset["data"].values
+    byte_count_list = unpacked_dataset["byte_count"].values
+
+    # The decompressed data in the shape of (epoch, n). Then reshape later.
+    decompressed_data = [
+        decompress(
+            packet_data[:byte_count],
+            compression_algorithm,
+        )
+        for (packet_data, byte_count) in zip(
+            binary_data_list, byte_count_list, strict=False
+        )
+    ]
+
+    # Look up collapse pattern using LUT table. This should return collapsed shape.
+    # For Lo species, it will be (1,)
+    collapsed_shape = get_collapse_pattern_shape(
+        sci_lut_data, view_tab_obj.sensor, view_tab_obj.collapse_table
+    )
+
+    # Reshape decompressed data to:
+    #   (num_packets, num_species, esa_steps, *collapsed_shape)
+    # where collapsed_shape is usually (1,) for Lo species.
+    num_packets = len(binary_data_list)
+    num_species = len(species_names)
+    esa_steps = constants.NUM_ESA_STEPS
+    species_data = np.array(decompressed_data).reshape(
+        num_packets, num_species, esa_steps, *collapsed_shape
+    )
+
+    # ========== Get Voltage Data from LUT ===========
+    # Use plan id and plan step to get voltage data's table_number in ESA sweep table.
+    # Voltage data is (128,)
+    esa_table_number = sci_lut_data["plan_tab"][f"({plan_id}, {plan_step})"][
+        "lo_stepping"
+    ]
+    voltage_data = sci_lut_data["esa_sweep_tab"][f"{esa_table_number}"]
+
+    # ========= Get Epoch Time Data ===========
+    # Epoch center time and delta
+    epoch_center, deltas = get_codice_epoch_time(
+        unpacked_dataset["acq_start_seconds"].values,
+        unpacked_dataset["acq_start_subseconds"].values,
+        unpacked_dataset["spin_period"].values,
+        view_tab_obj,
+    )
+
+    # ========== Create CDF Dataset with Metadata ===========
+    cdf_attrs = ImapCdfAttributes()
+    cdf_attrs.add_instrument_global_attrs("codice")
+    cdf_attrs.add_instrument_variable_attrs("codice", "l1a")
+
+    l1a_dataset = xr.Dataset(
+        coords={
+            "epoch": xr.DataArray(
+                epoch_center,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes("epoch", check_schema=False),
+            ),
+            "epoch_delta_minus": xr.DataArray(
+                deltas,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "epoch_delta_minus", check_schema=False
+                ),
+            ),
+            "epoch_delta_plus": xr.DataArray(
+                deltas,
+                dims=("epoch",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "epoch_delta_plus", check_schema=False
+                ),
+            ),
+            "esa_step": xr.DataArray(
+                np.arange(128),
+                dims=("esa_step",),
+                attrs=cdf_attrs.get_variable_attributes("esa_step", check_schema=False),
+            ),
+            "esa_step_label": xr.DataArray(
+                np.arange(128).astype(str),
+                dims=("esa_step",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "esa_step_label", check_schema=False
+                ),
+            ),
+            "k_factor": xr.DataArray(
+                np.array([constants.K_FACTOR]),
+                dims=("k_factor",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "k_factor_attrs", check_schema=False
+                ),
+            ),
+            "spin_sector": xr.DataArray(
+                np.array([0], dtype=np.uint8),
+                dims=("spin_sector",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "spin_sector", check_schema=False
+                ),
+            ),
+            "spin_sector_label": xr.DataArray(
+                np.array(["0"]).astype(str),
+                dims=("spin_sector",),
+                attrs=cdf_attrs.get_variable_attributes(
+                    "spin_sector_label", check_schema=False
+                ),
+            ),
+        },
+        attrs=cdf_attrs.get_global_attributes(logical_source_id),
+    )
+    # Add first few unique variables
+    l1a_dataset["spin_period"] = xr.DataArray(
+        unpacked_dataset["spin_period"].values * constants.SPIN_PERIOD_CONVERSION,
+        dims=("epoch",),
+        attrs=cdf_attrs.get_variable_attributes("spin_period"),
+    )
+    l1a_dataset["k_factor"] = xr.DataArray(
+        np.array([constants.K_FACTOR]),
+        dims=("k_factor",),
+        attrs=cdf_attrs.get_variable_attributes("k_factor_attrs", check_schema=False),
+    )
+    l1a_dataset["voltage_table"] = xr.DataArray(
+        np.array(voltage_data),
+        dims=("esa_step",),
+        attrs=cdf_attrs.get_variable_attributes("voltage_table", check_schema=False),
+    )
+    l1a_dataset["data_quality"] = xr.DataArray(
+        unpacked_dataset["suspect"].values,
+        dims=("epoch",),
+        attrs=cdf_attrs.get_variable_attributes("data_quality"),
+    )
+    l1a_dataset["acquisition_time_per_step"] = xr.DataArray(
+        calculate_acq_time_per_step(sci_lut_data["lo_stepping_tab"]),
+        dims=("esa_step",),
+        attrs=cdf_attrs.get_variable_attributes(
+            "acquisition_time_per_step", check_schema=False
+        ),
+    )
+
+    # Carry over these variables from unpacked data to l1a_dataset
+    l1a_carryover_vars = [
+        "sw_bias_gain_mode",
+        "st_bias_gain_mode",
+        "rgfo_half_spin",
+        "nso_half_spin",
+    ]
+    # Loop through them since we need to set their attrs too
+    for var in l1a_carryover_vars:
+        l1a_dataset[var] = xr.DataArray(
+            unpacked_dataset[var].values,
+            dims=("epoch",),
+            attrs=cdf_attrs.get_variable_attributes(var),
+        )
+
+    # Finally, add species data variables and their uncertainties
+    for idx, species in enumerate(species_names):
+        if view_tab_obj.apid == CODICEAPID.COD_LO_SW_SPECIES_COUNTS and species in [
+            "heplus",
+            "cnoplus",
+        ]:
+            species_attrs = cdf_attrs.get_variable_attributes("lo-pui-species-attrs")
+            unc_attrs = cdf_attrs.get_variable_attributes("lo-pui-species-unc-attrs")
+        else:
+            species_attrs = cdf_attrs.get_variable_attributes("lo-species-attrs")
+            unc_attrs = cdf_attrs.get_variable_attributes("lo-species-unc-attrs")
+
+        direction = (
+            "Sunward"
+            if view_tab_obj.apid == CODICEAPID.COD_LO_SW_SPECIES_COUNTS
+            else "Non-Sunward"
+        )
+        # Replace {species} and {direction} in attrs
+        species_attrs["CATDESC"] = species_attrs["CATDESC"].format(
+            species=species, direction=direction
+        )
+        species_attrs["FIELDNAM"] = species_attrs["FIELDNAM"].format(
+            species=species, direction=direction
+        )
+        l1a_dataset[species] = xr.DataArray(
+            species_data[:, idx, :, :],
+            dims=("epoch", "esa_step", "spin_sector"),
+            attrs=species_attrs,
+        )
+        # Uncertainty data
+        unc_attrs["CATDESC"] = unc_attrs["CATDESC"].format(
+            species=species, direction=direction
+        )
+        unc_attrs["FIELDNAM"] = unc_attrs["FIELDNAM"].format(
+            species=species, direction=direction
+        )
+        l1a_dataset[f"unc_{species}"] = xr.DataArray(
+            np.sqrt(species_data[:, idx, :, :]),
+            dims=("epoch", "esa_step", "spin_sector"),
+            attrs=unc_attrs,
+        )
+
+    return l1a_dataset