Connection
Base URL, credentials and a reachable server.
Product
Product features · Compensated floating-point sums
01
Trace the running total and correction beside cancellation fixtures so a small addend is not judged from the rounded total alone.
Selection
A model available through the configured endpoint.
Capabilities
Context limits and tool support are determined by the model and server.
Compensated floating-point sums
● ● ●SumHarrier●
A concrete starting point · Compensated floating-point sums ×
def compensated_sum(values):
values = tuple(values)
if len(values) > 1000 or any(type(v) not in (int, float) or
not -1e100 <= v <= 1e100 for v in values):
raise ValueError("values.range")
total = correction = 0.0
for value in values:
value = float(value)
next_total = total + value
correction += ((total - next_total) + value if abs(total) >= abs(value)
else (value - next_total) + total)
total = next_total
return total + correction✳Sum at most 1000 finite numbers of magnitude at most 1e100 with a compensation term.
✓ ChangesReview before applying ↗
Illustrative workspace