Problem

On an NXP Semiconductors MCU such as the LPC55S69, a UART baud rate is derived from a system clock and an integer divider. Given a system clock, an oversampling factor, and a list of target baud rates, write a function that computes the best divider for each target baud and returns the resulting actual baud rate and percentage error.

Use the formula:

actual_baud = system_clock / (oversample * divider)

For each target baud rate, choose the positive integer divider that minimizes absolute percentage error. If two dividers produce the same error, return the smaller divider.

Formal Specification

• Input:
  • system_clock: integer clock frequency in Hz
  • oversample: integer oversampling factor
  • targets: list of integer target baud rates
• Output:
  • A list of lists, where each result is: [target_baud, divider, actual_baud, error_percent]
  • actual_baud and error_percent should be rounded to 2 decimal places.

Examples

Example 1 Input: system_clock = 12000000, oversample = 16, targets = [9600] Output: [[9600, 78, 9615.38, 0.16]] Explanation: Divider 78 gives 12000000 / (16 * 78) = 9615.38, which is the closest achievable baud.

Example 2 Input: system_clock = 48000000, oversample = 16, targets = [115200, 1000000] Output: [[115200, 26, 115384.62, 0.16], [1000000, 3, 1000000.0, 0.0]] Explanation: Divider 26 is best for 115200, and divider 3 gives an exact 1,000,000 baud.

Constraints

• 1 <= len(targets) <= 10^4
• 1 <= system_clock <= 10^9
• 1 <= oversample <= 256
• 1 <= targets[i] <= 10^7
• Divider must be a positive integer.