Sigma Computing Onsite Second Round Interview Experience (USA)

Context The interview focuses on a Spreadsheet class initialized with specific dimensions ($M$ rows, $N$ columns). The class provides the following pre-built methods: * `set_cell(row, col, value

Context The interview focuses on a Spreadsheet class initialized with specific dimensions ($M$ rows, $N$ columns). The class provides the following pre-built methods: * set_cell(row, col, value): Assigns a value or formula to a specific cell. * get_cell_value(row, col): A stubbed function that needs to be implemented. * print_sheet(): Iterates through the entire grid to print resolved values. * parse_formula(formula): A utility function that breaks down a formula string into a list of operand coordinates. * op(a, b): A method that performs an operation (such as addition) between two integers and increments a global execution counter. Part 1: Formula Evaluation The first task requires implementing get_cell_value(row, col). A cell's content is either a direct integer or a formula string in the format =(row1, col1)@(row2, col2)@(row3, col3).... * The implementation must distinguish between raw values and formulas. * If a formula is present, parse_formula must be used to retrieve dependent coordinates (e.g., [(row1, col1), (row2, col2), ...]). * The values from these coordinates must be resolved recursively and combined using the op function to return the final integer result. Part 2: Performance Optimization The second task addresses computational efficiency. Since every call to op increments a counter, the objective is to minimize these calls during redundant operations. For example, if print_sheet() is called twice consecutively without any data changes, the system should not re-calculate the formulas. The solution requires modifying the Spreadsheet class to implement a caching or memoization strategy. This ensures that the op function is only invoked when necessary (i.e., when dependencies have changed), keeping the execution counter as low as possible.