Fermat MCP

import pytest # Test data MATRIX_2X2 = [[1, 2], [3, 4]] # Test data MATRIX_2X2 = [[1, 2], [3, 4]] MATRIX_3X3 = [[4, 1, 2], [1, 5, 3], [2, 3, 6]] VECTOR_1D = [1, 2, 3, 4, 5] VECTOR_2D = [[1, 2, 3], [4, 5, 6]] @pytest.mark.asyncio async def test_numerical_operations(client): """Test various numerical operations using the FastAPI test client.""" # Test create_array response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "create_array", "a": [1, 2, 3]}, ) assert response.data == [1, 2, 3] # Test zeros response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "zeros", "shape": [2, 3]} ) assert response.data == [[0.0, 0.0, 0.0], [0.0, 0.0, 0.0]] # Test ones response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "ones", "shape": [2, 2]} ) assert response.data == [[1.0, 1.0], [1.0, 1.0]] # Test full response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "full", "shape": [2, 2], "fill_value": 5}, ) assert response.data == [[5.0, 5.0], [5.0, 5.0]] # Test reshape response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "reshape", "a": [1, 2, 3, 4], "new_shape": [2, 2]}, ) assert response.data == [[1, 2], [3, 4]] # Test flatten response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "flatten", "a": [[1, 2], [3, 4]]}, ) assert response.data == [1, 2, 3, 4] # Test transpose response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "transpose", "a": [[1, 2], [3, 4]]}, ) assert response.data == [[1, 3], [2, 4]] # Test add response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "add", "a": [1, 2, 3], "b": [4, 5, 6]}, ) assert response.data == [5, 7, 9] # Test multiply response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "multiply", "a": [1, 2, 3], "b": [4, 5, 6]}, ) assert response.data == [4, 10, 18] # Test power response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "power", "a": [2, 3, 4], "b": [2, 3, 0.5]}, ) assert response.data == [4.0, 27.0, 2.0] # Test mean response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "mean", "a": [1, 2, 3, 4, 5]} ) assert response.data == 3.0 # Test std response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "std", "a": [1, 2, 3, 4, 5]} ) assert abs(response.data - 1.4142) < 0.0001 # Test dot product response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "dot", "a": [1, 2, 3], "b": [4, 5, 6]}, ) assert response.data == 32.0 # Test matrix multiplication response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "matmul", "a": [[1, 2], [3, 4]], "b": [[5, 6], [7, 8]]}, ) assert response.data == [[19.0, 22.0], [43.0, 50.0]] # Test determinant response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "det", "a": [[4, 7], [2, 6]]} ) assert response.data == pytest.approx(10.0) # Test matrix inverse response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "inv", "a": [[4, 7], [2, 6]]} ) expected = [[0.6, -0.7], [-0.2, 0.4]] for r, e in zip(response.data, expected): for a, b in zip(r, e): assert abs(a - b) < 1e-10 # Test eigenvalues and eigenvectors response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "eig", "a": [[1, -1], [1, 1]]}, ) assert len(response.data["eigenvalues"]) == 2 assert len(response.data["eigenvectors"]) == 2 assert len(response.data["eigenvectors"][0]) == 2 # Test eigenvalues only response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "eigenvals", "a": [[4, 1], [2, 3]]}, ) assert len(response.data) == 2 assert all(isinstance(x, (float, complex)) for x in response.data) # Test solving linear system response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "solve", "a": [[3, 1], [1, 2]], "b": [9, 8]}, ) expected = [2.0, 3.0] for r, e in zip(response.data, expected): assert abs(r - e) < 1e-10 # Test SVD response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "svd", "a": [[1, 2], [3, 4], [5, 6]]}, ) assert len(response.data["U"]) == 3 assert len(response.data["S"]) == 2 assert len(response.data["Vt"]) == 2 # Test invalid operation with pytest.raises(Exception) as exc_info: await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "invalid_operation", "a": [1, 2, 3]}, ) assert "Unknown operation" in str(exc_info.value) # Test empty array - expect an error with pytest.raises(Exception) as exc_info: await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "mean", "a": []} ) assert "Invalid structured content" in str(exc_info.value) # Test matrix multiplication dimension mismatch with pytest.raises(Exception) as exc_info: await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "matmul", "a": [[1, 2], [3, 4]], "b": [1, 2, 3]}, ) assert "mismatch in its core dimension" in str(exc_info.value) # Test real number operations response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "add", "a": [1, 2, 3], "b": [4, 5, 6]}, ) assert response.data == [5, 7, 9] # Test matrix multiplication with real numbers response = await client.call_tool( "numpy_mcp_numerical_operation", {"operation": "matmul", "a": [[1, 2], [3, 4]], "b": [[5, 6], [7, 8]]}, ) assert response.data == [[19.0, 22.0], [43.0, 50.0]] if __name__ == "__main__": pytest.main([__file__, "-v"]) # Test matrix multiplication with real numbers