solarwindpy 0.0.1.dev0__py3-none-any.whl → 0.1.1__py3-none-any.whl

tests/plotting/test_agg_plot.py

@@ -0,0 +1,602 @@
+#!/usr/bin/env python
+"""Tests for solarwindpy.plotting.agg_plot module.
+
+This module provides comprehensive test coverage for the AggPlot abstract base class
+used for aggregated plotting functionality including histograms and heatmaps.
+"""
+
+import pytest
+import numpy as np
+import pandas as pd
+from unittest.mock import patch, MagicMock
+
+from solarwindpy.plotting.agg_plot import AggPlot
+
+
+class ConcreteAggPlot(AggPlot):
+    """Concrete implementation of AggPlot for testing abstract functionality."""
+
+    def __init__(self, data=None, gb_axes=None, axnorm=None):
+        if data is None:
+            # Create synthetic test data
+            np.random.seed(42)
+            n = 100
+            data = pd.DataFrame(
+                {
+                    "x": np.random.normal(5, 2, n),
+                    "y": np.random.normal(10, 3, n),
+                    "z": np.random.normal(1, 0.5, n),
+                }
+            )
+        self._data = data
+        self._clip = False
+        self._gb_axes_list = gb_axes or ["x", "y"]
+        self._axnorm = axnorm or "none"
+        self._clim = (None, None)
+        self._alim = (None, None)
+
+        # Call super().__init__() after setting up required attributes
+        super().__init__()
+
+        # Initialize bins/intervals
+        self.calc_bins_intervals(nbins=10)
+        self.make_cut()
+
+    @property
+    def _gb_axes(self):
+        """The axes over which the groupby aggregation takes place."""
+        return self._gb_axes_list
+
+    def make_plot(self):
+        """Minimal implementation for testing."""
+        return "agg_plot_created"
+
+    def set_data(self, data):
+        """Minimal implementation for testing."""
+        self._data = data
+
+    def set_path(self, new, add_scale=False):
+        """Concrete implementation of set_path for testing."""
+        path, x, y, z, scale_info = super().set_path(new, add_scale)
+        self._path = path
+        return path, x, y, z, scale_info
+
+    def set_axnorm(self, new):
+        """Concrete implementation for testing."""
+        self._axnorm = new
+
+
+class TestAggPlotProperties:
+    """Test the AggPlot properties."""
+
+    def test_edges_property(self):
+        """Test that edges property constructs correct bin-edge arrays."""
+        agg_plot = ConcreteAggPlot()
+        edges = agg_plot.edges
+
+        assert isinstance(edges, dict)
+        assert "x" in edges
+        assert "y" in edges
+
+        # Check that edges are numeric and increasing
+        for axis, edge_array in edges.items():
+            assert hasattr(edge_array, "__iter__")
+            edge_list = list(edge_array)
+            assert len(edge_list) > 1
+            # Edges should be sorted
+            assert edge_list == sorted(edge_list)
+
+    def test_categoricals_property(self):
+        """Test that categoricals property returns categorical bins mapping."""
+        agg_plot = ConcreteAggPlot()
+        categoricals = agg_plot.categoricals
+
+        assert isinstance(categoricals, dict)
+        assert "x" in categoricals
+        assert "y" in categoricals
+
+        for axis, cat_index in categoricals.items():
+            assert hasattr(cat_index, "__iter__")
+            # Should contain interval objects
+            assert len(cat_index) > 0
+
+    def test_intervals_property(self):
+        """Test that intervals property returns correct IntervalIndex objects."""
+        agg_plot = ConcreteAggPlot()
+        intervals = agg_plot.intervals
+
+        assert isinstance(intervals, dict)
+        assert "x" in intervals
+        assert "y" in intervals
+
+        for axis, interval_index in intervals.items():
+            assert isinstance(interval_index, pd.IntervalIndex)
+            assert len(interval_index) > 0
+            # All intervals should be valid
+            for interval in interval_index:
+                assert hasattr(interval, "left")
+                assert hasattr(interval, "right")
+                assert interval.left <= interval.right
+
+    def test_cut_property(self):
+        """Test that cut property returns the internal _cut DataFrame."""
+        agg_plot = ConcreteAggPlot()
+        cut = agg_plot.cut
+
+        assert isinstance(cut, pd.DataFrame)
+        assert "x" in cut.columns
+        assert "y" in cut.columns
+        assert len(cut) == len(agg_plot.data)
+        assert cut is agg_plot._cut
+
+    def test_clim_property(self):
+        """Test that clim property returns the internal _clim tuple."""
+        agg_plot = ConcreteAggPlot()
+
+        # Test default
+        assert agg_plot.clim == (None, None)
+        assert agg_plot.clim is agg_plot._clim
+
+        # Test after setting
+        agg_plot.set_clim(5, 100)
+        assert agg_plot.clim == (5, 100)
+
+    def test_alim_property(self):
+        """Test that alim property returns the internal _alim tuple."""
+        agg_plot = ConcreteAggPlot()
+
+        # Test default
+        assert agg_plot.alim == (None, None)
+        assert agg_plot.alim is agg_plot._alim
+
+        # Test after setting
+        agg_plot.set_alim(0.1, 0.9)
+        assert agg_plot.alim == (0.1, 0.9)
+
+    def test_agg_axes_property(self):
+        """Test that agg_axes returns the correct aggregation column."""
+        # Test with 2D groupby axes (x, y), z should be agg axis
+        agg_plot = ConcreteAggPlot()
+        agg_axis = agg_plot.agg_axes
+
+        assert agg_axis == "z"  # The column not in _gb_axes
+
+        # Test with 1D groupby axes, need exactly one non-groupby column
+        data = pd.DataFrame(
+            {"x": [1, 2, 3], "z": [7, 8, 9]}  # Only one non-groupby column
+        )
+        agg_plot_1d = ConcreteAggPlot(data=data, gb_axes=["x"])
+
+        # Should return the single non-groupby column
+        agg_axis_1d = agg_plot_1d.agg_axes
+        assert agg_axis_1d == "z"
+
+    def test_joint_property(self):
+        """Test that joint returns a Series with a MultiIndex."""
+        agg_plot = ConcreteAggPlot()
+        joint = agg_plot.joint
+
+        assert isinstance(joint, pd.Series)
+        assert isinstance(joint.index, pd.MultiIndex)
+
+        # Index should have levels for each groupby axis
+        assert len(joint.index.levels) == len(agg_plot._gb_axes)
+
+        # Should contain data from the aggregation axis
+        assert joint.name == agg_plot.agg_axes
+
+    def test_grouped_property(self):
+        """Test that grouped returns a GroupBy on the correct axes."""
+        agg_plot = ConcreteAggPlot()
+        grouped = agg_plot.grouped
+
+        assert hasattr(grouped, "agg")  # Should be a GroupBy object
+        assert hasattr(grouped, "groups")
+
+        # Should be grouped by the gb_axes - check using keys instead of deprecated grouper
+        assert hasattr(grouped, "keys")  # GroupBy object should have keys method
+
+    def test_axnorm_property(self):
+        """Test that axnorm returns the internal _axnorm value."""
+        agg_plot = ConcreteAggPlot(axnorm="row")
+
+        assert agg_plot.axnorm == "row"
+        assert agg_plot.axnorm is agg_plot._axnorm
+
+        # Test setting new value
+        agg_plot.set_axnorm("column")
+        assert agg_plot.axnorm == "column"
+
+
+class TestAggPlotClipData:
+    """Test the clip_data static method.
+
+    Note: The current implementation uses deprecated pandas methods (clip_lower, clip_upper).
+    These tests verify the method signature but expect AttributeError for deprecated methods.
+    """
+
+    def test_clip_data_series_lower(self):
+        """Test clip_data with Series and 'l' (lower) clipping."""
+        data = pd.Series([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
+
+        # The implementation uses deprecated clip_lower method
+        with pytest.raises(
+            AttributeError, match="'Series' object has no attribute 'clip_lower'"
+        ):
+            AggPlot.clip_data(data, "l")
+
+    def test_clip_data_series_upper(self):
+        """Test clip_data with Series and 'u' (upper) clipping."""
+        data = pd.Series([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
+
+        # The implementation uses deprecated clip_upper method
+        with pytest.raises(
+            AttributeError, match="'Series' object has no attribute 'clip_upper'"
+        ):
+            AggPlot.clip_data(data, "u")
+
+    def test_clip_data_series_both(self):
+        """Test clip_data with Series and both upper/lower clipping."""
+        data = pd.Series([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
+
+        result = AggPlot.clip_data(
+            data, "both"
+        )  # Any string not starting with 'l' or 'u'
+
+        assert isinstance(result, pd.Series)
+        assert len(result) == len(data)
+        # Both upper and lower should be clipped
+        assert result.min() >= data.quantile(0.0001)
+        assert result.max() <= data.quantile(0.9999)
+
+    def test_clip_data_dataframe_lower(self):
+        """Test clip_data with DataFrame and 'l' clipping."""
+        data = pd.DataFrame({"a": [1, 2, 3, 4, 5], "b": [10, 20, 30, 40, 50]})
+
+        # The implementation uses deprecated clip_lower method
+        with pytest.raises(
+            AttributeError, match="'DataFrame' object has no attribute 'clip_lower'"
+        ):
+            AggPlot.clip_data(data, "l")
+
+    def test_clip_data_dataframe_upper(self):
+        """Test clip_data with DataFrame and 'u' clipping."""
+        data = pd.DataFrame({"a": [1, 2, 3, 4, 5], "b": [10, 20, 30, 40, 50]})
+
+        # The implementation uses deprecated clip_upper method
+        with pytest.raises(
+            AttributeError, match="'DataFrame' object has no attribute 'clip_upper'"
+        ):
+            AggPlot.clip_data(data, "u")
+
+    def test_clip_data_dataframe_both(self):
+        """Test clip_data with DataFrame and both clipping."""
+        data = pd.DataFrame({"a": [1, 2, 3, 4, 5], "b": [10, 20, 30, 40, 50]})
+
+        result = AggPlot.clip_data(data, "both")
+
+        assert isinstance(result, pd.DataFrame)
+        assert result.shape == data.shape
+        # Both bounds should be clipped column-wise
+        for col in data.columns:
+            assert result[col].min() >= data[col].quantile(0.0001)
+            assert result[col].max() <= data[col].quantile(0.9999)
+
+    def test_clip_data_unsupported_type_raises_error(self):
+        """Test that clip_data raises AttributeError on unsupported input types."""
+        unsupported_data = [1, 2, 3, 4, 5]  # List instead of Series/DataFrame
+
+        # The method first calls quantile(), which fails on lists
+        with pytest.raises(
+            AttributeError, match="'list' object has no attribute 'quantile'"
+        ):
+            AggPlot.clip_data(unsupported_data, "l")
+
+    def test_clip_data_case_insensitive(self):
+        """Test that clip_data handles case-insensitive mode strings."""
+        data = pd.Series([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
+
+        # Test uppercase - both should raise same error for deprecated methods
+        with pytest.raises(AttributeError):
+            AggPlot.clip_data(data, "L")
+
+        with pytest.raises(AttributeError):
+            AggPlot.clip_data(data, "U")
+
+
+class TestAggPlotSetMethods:
+    """Test the set_* methods in AggPlot."""
+
+    def test_set_clim_valid_numbers(self):
+        """Test set_clim with valid number inputs."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.set_clim(2, 10)
+        assert agg_plot._clim == (2, 10)
+
+        # Test with floats
+        agg_plot.set_clim(1.5, 9.5)
+        assert agg_plot._clim == (1.5, 9.5)
+
+    def test_set_clim_none_values(self):
+        """Test set_clim with None values."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.set_clim(None, 10)
+        assert agg_plot._clim == (None, 10)
+
+        agg_plot.set_clim(2, None)
+        assert agg_plot._clim == (2, None)
+
+        agg_plot.set_clim(None, None)
+        assert agg_plot._clim == (None, None)
+
+    def test_set_clim_invalid_types(self):
+        """Test set_clim with invalid types raises AssertionError."""
+        agg_plot = ConcreteAggPlot()
+
+        with pytest.raises(AssertionError):
+            agg_plot.set_clim("invalid", 10)
+
+        with pytest.raises(AssertionError):
+            agg_plot.set_clim(2, "invalid")
+
+        with pytest.raises(AssertionError):
+            agg_plot.set_clim([1, 2], 10)
+
+    def test_set_alim_valid_numbers(self):
+        """Test set_alim with valid number inputs."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.set_alim(0.1, 0.9)
+        assert agg_plot._alim == (0.1, 0.9)
+
+        # Test with integers
+        agg_plot.set_alim(1, 100)
+        assert agg_plot._alim == (1, 100)
+
+    def test_set_alim_none_values(self):
+        """Test set_alim with None values."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.set_alim(None, 0.9)
+        assert agg_plot._alim == (None, 0.9)
+
+        agg_plot.set_alim(0.1, None)
+        assert agg_plot._alim == (0.1, None)
+
+    def test_set_alim_invalid_types(self):
+        """Test set_alim with invalid types raises AssertionError."""
+        agg_plot = ConcreteAggPlot()
+
+        with pytest.raises(AssertionError):
+            agg_plot.set_alim("invalid", 0.9)
+
+        with pytest.raises(AssertionError):
+            agg_plot.set_alim(0.1, [0.9])
+
+
+class TestAggPlotBinCalculation:
+    """Test bin calculation and cutting functionality."""
+
+    def test_calc_bins_intervals_default(self):
+        """Test calc_bins_intervals with default parameters."""
+        agg_plot = ConcreteAggPlot()
+
+        # Should have calculated intervals and categoricals
+        assert hasattr(agg_plot, "_categoricals")
+        assert len(agg_plot._categoricals) == len(agg_plot._gb_axes)
+
+        for axis, intervals in agg_plot._categoricals:
+            assert axis in agg_plot._gb_axes
+            assert len(intervals) > 0
+
+    def test_calc_bins_intervals_custom_nbins(self):
+        """Test calc_bins_intervals with custom number of bins."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.calc_bins_intervals(nbins=5)
+
+        # Check that we get approximately the requested number of bins
+        for axis, intervals in agg_plot._categoricals:
+            # Should be roughly 5 bins (may vary due to data distribution)
+            assert 3 <= len(intervals) <= 7
+
+    def test_calc_bins_intervals_different_nbins_per_axis(self):
+        """Test calc_bins_intervals with different bins per axis."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.calc_bins_intervals(nbins=[5, 8])  # 5 for x, 8 for y
+
+        intervals_dict = dict(agg_plot._categoricals)
+        x_intervals = intervals_dict["x"]
+        y_intervals = intervals_dict["y"]
+
+        # x should have ~5 bins, y should have ~8 bins
+        assert 3 <= len(x_intervals) <= 7
+        assert 6 <= len(y_intervals) <= 10
+
+    def test_calc_bins_intervals_with_precision(self):
+        """Test calc_bins_intervals with custom precision."""
+        agg_plot = ConcreteAggPlot()
+
+        agg_plot.calc_bins_intervals(nbins=5, precision=2)
+
+        # Check that intervals exist (specific precision testing would require
+        # examining the actual bin edges, which is implementation dependent)
+        assert hasattr(agg_plot, "_categoricals")
+        assert len(agg_plot._categoricals) > 0
+
+    def test_make_cut_creates_cut_dataframe(self):
+        """Test that make_cut creates the _cut DataFrame."""
+        agg_plot = ConcreteAggPlot()
+
+        # make_cut should have been called in __init__
+        assert hasattr(agg_plot, "_cut")
+        assert isinstance(agg_plot._cut, pd.DataFrame)
+        assert len(agg_plot._cut) == len(agg_plot.data)
+        assert list(agg_plot._cut.columns) == agg_plot._gb_axes
+
+    def test_make_cut_with_clipping(self):
+        """Test make_cut with data clipping enabled."""
+        agg_plot = ConcreteAggPlot()
+        agg_plot._clip = "l"  # Enable lower clipping
+
+        # The current implementation uses deprecated pandas methods, so this will fail
+        with pytest.raises(
+            AttributeError, match="'Series' object has no attribute 'clip_lower'"
+        ):
+            agg_plot.make_cut()
+
+
+class TestAggPlotAggregation:
+    """Test data aggregation functionality."""
+
+    def test_agg_default_behavior(self):
+        """Test agg method with default parameters."""
+        agg_plot = ConcreteAggPlot()
+
+        result = agg_plot.agg()
+
+        assert isinstance(result, pd.Series)
+        assert isinstance(result.index, pd.MultiIndex)
+        assert len(result.index.levels) == len(agg_plot._gb_axes)
+
+    def test_agg_with_count_function(self):
+        """Test agg method with count function."""
+        agg_plot = ConcreteAggPlot()
+
+        result = agg_plot.agg(fcn="count")
+
+        assert isinstance(result, pd.Series)
+        assert all(result.dropna() >= 0)  # Counts should be non-negative
+
+    def test_agg_with_mean_function(self):
+        """Test agg method with mean function."""
+        agg_plot = ConcreteAggPlot()
+
+        result = agg_plot.agg(fcn="mean")
+
+        assert isinstance(result, pd.Series)
+        # Values should be reasonable (within data range)
+        data_min = agg_plot.data[agg_plot.agg_axes].min()
+        data_max = agg_plot.data[agg_plot.agg_axes].max()
+        result_clean = result.dropna()
+        if len(result_clean) > 0:
+            assert (
+                result_clean.min() >= data_min - 1e-10
+            )  # Allow for floating point error
+            assert result_clean.max() <= data_max + 1e-10
+
+    def test_agg_with_clim_filtering(self):
+        """Test agg method with count limits applied."""
+        agg_plot = ConcreteAggPlot()
+        agg_plot.set_clim(2, None)  # Require at least 2 counts per bin
+
+        result = agg_plot.agg()
+
+        assert isinstance(result, pd.Series)
+        # Some bins may be filtered out due to low counts
+        # Exact behavior depends on data distribution
+
+    def test_agg_with_kwargs(self):
+        """Test agg method with additional kwargs."""
+        agg_plot = ConcreteAggPlot()
+
+        # Should not raise an error when passing additional aggregation kwargs
+        result = agg_plot.agg(fcn="mean")
+        assert isinstance(result, pd.Series)
+
+
+class TestAggPlotUtilityMethods:
+    """Test utility methods in AggPlot."""
+
+    def test_get_plotted_data_boolean_series(self):
+        """Test get_plotted_data_boolean_series method."""
+        agg_plot = ConcreteAggPlot()
+
+        boolean_series = agg_plot.get_plotted_data_boolean_series()
+
+        assert isinstance(boolean_series, pd.Series)
+        assert boolean_series.dtype == bool
+        assert len(boolean_series) == len(agg_plot.data)
+        # All values should be boolean
+        assert boolean_series.isin([True, False]).all()
+
+    def test_get_subset_above_threshold(self):
+        """Test get_subset_above_threshold method."""
+        agg_plot = ConcreteAggPlot()
+
+        subset, boolean_mask = agg_plot.get_subset_above_threshold(threshold=1)
+
+        assert isinstance(subset, pd.DataFrame)
+        assert isinstance(boolean_mask, pd.Series)
+        assert boolean_mask.dtype == bool
+        assert len(boolean_mask) == len(agg_plot.data)
+        assert len(subset) == boolean_mask.sum()
+
+    def test_get_subset_above_threshold_with_custom_function(self):
+        """Test get_subset_above_threshold with custom aggregation function."""
+        agg_plot = ConcreteAggPlot()
+
+        subset, boolean_mask = agg_plot.get_subset_above_threshold(
+            threshold=0.5, fcn="mean"
+        )
+
+        assert isinstance(subset, pd.DataFrame)
+        assert isinstance(boolean_mask, pd.Series)
+
+
+class TestAggPlotEdgeCases:
+    """Test edge cases and error conditions."""
+
+    def test_agg_plot_with_minimal_data(self):
+        """Test AggPlot with minimal data."""
+        minimal_data = pd.DataFrame({"x": [1, 2], "y": [3, 4], "z": [5, 6]})
+
+        agg_plot = ConcreteAggPlot(data=minimal_data)
+
+        # Should still work with minimal data
+        assert len(agg_plot.data) == 2
+        assert hasattr(agg_plot, "_cut")
+
+    def test_agg_plot_with_nan_data(self):
+        """Test AggPlot handling of NaN values."""
+        data_with_nans = pd.DataFrame(
+            {"x": [1, 2, np.nan, 4], "y": [5, np.nan, 7, 8], "z": [9, 10, 11, np.nan]}
+        )
+
+        agg_plot = ConcreteAggPlot(data=data_with_nans)
+
+        # Should handle NaN values gracefully
+        assert len(agg_plot.data) == 4
+        # Aggregation should work despite NaNs
+        result = agg_plot.agg()
+        assert isinstance(result, pd.Series)
+
+    def test_abstract_properties_and_methods(self):
+        """Test that abstract properties and methods are properly implemented."""
+        agg_plot = ConcreteAggPlot()
+
+        # _gb_axes should be implemented
+        assert hasattr(agg_plot, "_gb_axes")
+        assert isinstance(agg_plot._gb_axes, list)
+
+        # set_axnorm should be implemented
+        agg_plot.set_axnorm("test")
+        assert agg_plot._axnorm == "test"
+
+    def test_logger_functionality(self):
+        """Test that logging functionality works."""
+        agg_plot = ConcreteAggPlot()
+
+        # Should have a logger
+        assert hasattr(agg_plot, "logger")
+        logger = agg_plot.logger
+
+        # Logger should have proper name
+        assert "AggPlot" in logger.name or "ConcreteAggPlot" in logger.name
+
+
+if __name__ == "__main__":
+    pytest.main([__file__])