WEBVTT

00:00:00.116 --> 00:00:02.281
With a fifty-percent starting guess, this

00:00:02.294 --> 00:00:04.665
financial library returned thirteen point

00:00:04.678 --> 00:00:06.353
five eight percent per period for

00:00:06.366 --> 00:00:07.255
a synthetic loan.

00:00:08.036 --> 00:00:09.299
Two hundred thousand borrowed.

00:00:09.776 --> 00:00:11.438
Three hundred sixty payments of one

00:00:11.451 --> 00:00:11.980
thousand.

00:00:12.405 --> 00:00:14.725
Direct discounting gives about zero point

00:00:14.828 --> 00:00:17.031
three six six percent per period.

00:00:17.783 --> 00:00:19.741
The array solver had not converged.

00:00:20.167 --> 00:00:22.332
Its final check ignored large negative

00:00:22.344 --> 00:00:22.899
steps.

00:00:23.337 --> 00:00:25.747
The correction returns failure here, and

00:00:25.759 --> 00:00:28.504
a second fix stabilizes calculations around

00:00:28.517 --> 00:00:28.968
zero.

00:00:29.708 --> 00:00:31.563
The corrected Rust code passed two

00:00:31.576 --> 00:00:33.922
hundred forty-two tests and four hundred

00:00:33.935 --> 00:00:34.978
nine extra checks.

00:00:35.417 --> 00:00:37.968
The report includes code and limitations.

00:00:38.393 --> 00:00:40.017
No customer losses were measured.