added module for calibration data

src/conda_helpers/__init__.py

@@ -8,5 +8,6 @@ __version__ = "0.0.2"
 
 from sensospot_tools import select, split  # noqa: F401
 
+from .calibration import prepare_calibration_data  # noqa: F401
 from .linear_regression import linear_regression  # noqa: F401
 from .mbp import add_exposure_info, normalize  # noqa: F401

src/conda_helpers/calibration.py

@@ -0,0 +1,83 @@
+import pandas as pd
+
+EXPOSURE_ID = "Exposure.Id"
+EXPOSURE_CHANNEL = "Exposure.Channel"
+EXPOSURE_TIME = "Exposure.Time"
+
+CLASSIC_CALIBRATION_EXPOSURES = 4
+
+EXPOSURE_COLUMNS = [
+    EXPOSURE_ID,
+    EXPOSURE_CHANNEL,
+    EXPOSURE_TIME,
+]
+
+POS_TO_IGNORE = [
+    1,
+    2,
+    9,
+    10,
+    11,
+    12,
+    19,
+    20,
+    71,
+    72,
+    81,
+    82,
+    89,
+    90,
+    91,
+    92,
+    99,
+    100,
+]
+
+SIGNAL_SOURCE = "Signal.Source"
+SIGNAL_INTENSITY = "Signal.Intensity"
+SPOT_MEDIAN = "Spot.Median"
+BKG_MEDIAN = "Bkg.Median"
+SIGNAL_COLUMNS = [SPOT_MEDIAN, BKG_MEDIAN]
+
+
+def add_exposure_info(data: pd.DataFrame) -> pd.DataFrame:
+    if len(data[EXPOSURE_ID].unique()) == CLASSIC_CALIBRATION_EXPOSURES:
+        exposure_values = [
+            (1, "Cy3", 100),
+            (2, "Cy3", 10),
+            (3, "Cy5", 60),
+            (4, "Cy5", 6),
+        ]
+    else:
+        exposure_values = [
+            (1, "Cy3", 100),
+            (2, "Cy3", 30),
+            (3, "Cy3", 10),
+            (4, "Cy5", 60),
+            (5, "Cy5", 18),
+            (6, "Cy5", 6),
+        ]
+    exposure_df = pd.DataFrame(exposure_values, columns=EXPOSURE_COLUMNS)
+    return data.merge(exposure_df, on=EXPOSURE_ID)
+
+
+def remove_non_signal_positions(data: pd.DataFrame) -> pd.DataFrame:
+    selection = data["Pos.Id"].isin(POS_TO_IGNORE)
+    return data.loc[~selection].copy()
+
+
+def add_signal_columns(data: pd.DataFrame) -> pd.DataFrame:
+    def _generate_signal_columns():
+        for column in SIGNAL_COLUMNS:
+            copied = data.copy()
+            copied[SIGNAL_SOURCE] = column
+            copied[SIGNAL_INTENSITY] = data[column]
+            yield copied
+
+    return pd.concat(_generate_signal_columns()).reset_index(drop=True)
+
+
+def prepare_calibration_data(data: pd.DataFrame) -> pd.DataFrame:
+    with_exposures = add_exposure_info(data)
+    only_signals = remove_non_signal_positions(with_exposures)
+    return add_signal_columns(only_signals)

tests/test_calibration.py

@@ -0,0 +1,83 @@
+import pandas as pd
+import pytest
+
+
+def _generate_example_data(exposures):
+    values = [(i, i * 10, i * 100) for i in range(1, exposures + 1)]
+    columns = ["Exposure.Id", "Bkg.Median", "Spot.Median"]
+
+    return pd.DataFrame(values, columns=columns)
+
+
+@pytest.fixture()
+def classic_example_data():
+    return _generate_example_data(4)
+
+
+@pytest.fixture()
+def regression_example_data():
+    return _generate_example_data(6)
+
+
+def test_add_exposure_info_classic(classic_example_data):
+    from conda_helpers.calibration import add_exposure_info
+
+    result = add_exposure_info(classic_example_data)
+
+    assert list(result["Exposure.Channel"]) == ["Cy3"] * 2 + ["Cy5"] * 2
+    assert list(result["Exposure.Time"]) == [100, 10, 60, 6]
+
+
+def test_add_exposure_info_regression(regression_example_data):
+    from conda_helpers.calibration import add_exposure_info
+
+    result = add_exposure_info(regression_example_data)
+
+    assert list(result["Exposure.Channel"]) == ["Cy3"] * 3 + ["Cy5"] * 3
+    assert list(result["Exposure.Time"]) == [100, 30, 10, 60, 18, 6]
+
+
+def test_remove_non_signal_positions():
+    from conda_helpers.calibration import remove_non_signal_positions
+
+    data = pd.DataFrame(range(1, 12), columns=["Pos.Id"])
+
+    result = remove_non_signal_positions(data)
+
+    assert set(result["Pos.Id"]) == {3, 4, 5, 6, 7, 8}
+
+
+def test_add_signal_columns(classic_example_data):
+    from conda_helpers.calibration import add_signal_columns
+
+    result = add_signal_columns(classic_example_data)
+
+    assert len(result) == 2 * len(classic_example_data)
+
+    assert (
+        list(result["Signal.Source"])
+        == ["Spot.Median"] * 4 + ["Bkg.Median"] * 4
+    )
+    assert list(result["Signal.Intensity"]) == [
+        100,
+        200,
+        300,
+        400,
+        10,
+        20,
+        30,
+        40,
+    ]
+
+
+def test_prepare_calibration_data(classic_example_data):
+    from conda_helpers.calibration import prepare_calibration_data
+
+    classic_example_data["Pos.Id"] = 0
+    result = prepare_calibration_data(classic_example_data)
+
+    assert len(result) == 2 * len(classic_example_data)
+    assert "Exposure.Channel" in result.columns
+    assert "Exposure.Time" in result.columns
+    assert "Signal.Source" in result.columns
+    assert "Signal.Intensity" in result.columns

tests/test_conda_helpers.py

@@ -27,6 +27,7 @@ def test_api():
         add_exposure_info,  # noqa: F401
         linear_regression,  # noqa: F401
         normalize,  # noqa: F401
+        prepare_calibration_data,  # noqa: F401
         select,  # noqa: F401
         split,  # noqa: F401
     )