"""Tests for non-spatially referenced objects""" from __future__ import annotations from typing import TYPE_CHECKING import numpy as np import pytest import vtk import pyvista as pv from pyvista import examples if TYPE_CHECKING: from pytest_mock import MockerFixture try: import pandas as pd except ImportError: pd = None def test_table_init(tmpdir): """Save some delimited text to a file and read it""" filename = str(tmpdir.mkdir('tmpdir').join('tmp.csv')) nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) # Create from 2D array table = pv.Table(arrays) assert table.n_rows == nr assert table.n_columns == nc assert table.n_arrays == nc assert len(table.row_arrays) == nc for i in range(nc): assert np.allclose(arrays[:, i], table[i]) with pytest.raises(ValueError): # noqa: PT011 pv.Table(np.random.default_rng().random((100, 2, 3))) # Create from 1D array table = pv.Table(arrays[:, 0]) assert table.n_rows == nr assert table.n_columns == 1 assert len(table.row_arrays) == 1 assert np.allclose(arrays[:, 0], table[0]) # create from dictionary array_dict = {} for i in range(nc): array_dict[f'foo{i}'] = arrays[:, i].copy() table = pv.Table(array_dict) assert table.n_rows == nr assert table.n_columns == nc assert len(table.row_arrays) == nc for i in range(nc): assert np.allclose(arrays[:, i], table[f'foo{i}']) dataset = examples.load_hexbeam() array_dict = dict(dataset.point_data) table = pv.Table(array_dict) assert table.n_rows == dataset.n_points assert table.n_columns == len(array_dict) assert len(table.row_arrays) == len(array_dict) for name in table.keys(): assert np.allclose(dataset[name], table[name]) # Create from vtkTable object h = '\t'.join([f'a{i}' for i in range(nc)]) np.savetxt(filename, arrays, delimiter='\t', header=h, comments='') reader = vtk.vtkDelimitedTextReader() reader.SetFileName(filename) reader.DetectNumericColumnsOn() reader.SetFieldDelimiterCharacters('\t') reader.SetHaveHeaders(True) reader.Update() # Test init table = pv.Table(reader.GetOutput(), deep=True) assert isinstance(table, vtk.vtkTable) assert isinstance(table, pv.Table) table = pv.Table(reader.GetOutput(), deep=False) assert isinstance(table, vtk.vtkTable) assert isinstance(table, pv.Table) # Test wrap table = pv.wrap(reader.GetOutput()) assert isinstance(table, vtk.vtkTable) assert isinstance(table, pv.Table) assert table.n_rows == nr assert table.n_columns == nc assert len(table.row_arrays) == nc for i in range(nc): assert np.allclose(arrays[:, i], table[i]) with pytest.raises(TypeError): pv.Table('foo') def test_table_row_arrays(): nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) table = pv.Table() for i in range(nc): table[f'foo{i}'] = arrays[:, i] assert table.n_columns == nc assert table.n_rows == nr for i in range(nc): assert np.allclose(table[f'foo{i}'], arrays[:, i]) # Multi component table = pv.Table() table['multi'] = arrays assert table.n_columns == 1 assert table.n_rows == nr assert np.allclose(table[0], arrays) assert np.allclose(table['multi'], arrays) del table['multi'] assert table.n_columns == 0 dataset = examples.load_hexbeam() array_dict = dataset.point_data # Test dict methods table = pv.Table() table.update(array_dict) assert table.n_rows == dataset.n_points assert table.n_columns == len(array_dict) assert len(table.row_arrays) == len(array_dict) for name in table.keys(): assert np.allclose(dataset[name], table[name]) for i, array in enumerate(table.values()): name = table.keys()[i] assert np.allclose(dataset[name], array) for name, array in table.items(): assert np.allclose(dataset[name], array) n = table.n_arrays array = table.pop(table.keys()[0]) assert isinstance(array, np.ndarray) assert table.n_arrays == n - 1 array = table.get(table.keys()[0]) assert isinstance(array, np.ndarray) assert table.n_arrays == n - 1 del table[table.keys()[0]] assert table.n_arrays == n - 2 def test_table_row_np_bool(): n = 50 table = pv.Table() bool_arr = np.zeros(n, np.bool_) table.row_arrays['bool_arr'] = bool_arr bool_arr[:] = True assert table.row_arrays['bool_arr'].all() assert table._row_array('bool_arr').all() assert table._row_array('bool_arr').dtype == np.bool_ def test_table_row_uint8(): n = 50 table = pv.Table() arr = np.zeros(n, np.uint8) table.row_arrays['arr'] = arr arr[:] = np.arange(n) assert np.allclose(table.row_arrays['arr'], np.arange(n)) def test_table_repr(): nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) table = pv.Table(arrays) text = table._repr_html_() assert isinstance(text, str) text = table.__repr__() assert isinstance(text, str) text = table.__str__() assert isinstance(text, str) @pytest.mark.skipif(pd is None, reason='Requires Pandas') def test_table_pandas(): nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) df = pd.DataFrame() for i in range(nc): df[f'foo{i}'] = arrays[:, i].copy() table = pv.Table(df) assert table.n_rows == nr assert table.n_columns == nc for i in range(nc): assert np.allclose(table.row_arrays[f'foo{i}'], arrays[:, i]) assert df.equals(table.to_pandas()) def test_table_iter(): nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) table = pv.Table(arrays) for i, array in enumerate(table): assert np.allclose(array, arrays[:, i]) @pytest.mark.parametrize('preference', ['row', None]) def test_get_data_range_table(preference): nr, nc = 50, 3 arrays = np.random.default_rng().random((nr, nc)) table = pv.Table(arrays) nanmin, nanmax = ( table.get_data_range(preference=preference) if preference else table.get_data_range() ) assert nanmin == np.nanmin(arrays[:, 0]) assert nanmax == np.nanmax(arrays[:, 0]) def test_from_dict_raises(mocker: MockerFixture): m = mocker.MagicMock() m.ndim = 1 with pytest.raises( ValueError, match='Dictionary must contain only NumPy arrays with maximum of 2D.' ): pv.Table(dict(a=m))