Closed PatrickGreen123 closed 7 years ago
I think I have fixed this now
Hi there, this looks good. Could you please provide the test data set/ triangle that caused the issue. I't like to add a test to the build process, ensuring the issue doesn't arise again. Thanks
Hi Markus,
I can't give you the triangle that revealed the bug, but you can find attached a made up triangle which should reproduce the NAN issue.
The triangle is completely unrealistic but hopefully that doesn't matter to you.
Kind Regards,
Patrick
From: Markus Gesmann notifications@github.com Sent: 13 October 2017 08:54:03 To: mages/ChainLadder Cc: PatrickGreen123; State change Subject: Re: [mages/ChainLadder] Mack method standard errors NaN issue (#39)
Hi there, this looks good. Could you please provide the test data set/ triangle that caused the issue. I't like to add a test to the build process, ensuring the issue doesn't arise again. Thanks
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHubhttps://github.com/mages/ChainLadder/issues/39#issuecomment-336392204, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AcwmAEQVzuYIDrlOVl8Jn8Km6jypxqelks5sryUqgaJpZM4P3f0o.
Hi Patrick, a made up triangle is exactly, what I am looking for. Unfortunately, it didn't seem to attached. Best, it you load it into R and paste the output of
dput(triangle)
Thanks Markus
Try this link:
https://drive.google.com/file/d/0B605tTCPt_5sT2JvRGswUW1STVE/view?usp=drivesdk
If that doesn't work then I'll try your suggestion
---- Markus Gesmann wrote ----
Hi Patrick, a made up triangle is exactly, what I am looking for. Unfortunately, it didn't seem to attached. Best, it you load into R and paste the output of
dput(triangle)
Thanks Markus
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHubhttps://github.com/mages/ChainLadder/issues/39#issuecomment-337373180, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AcwmAIrG-a-bVZSx3MUHNK_O-1lKm7r5ks5stRgsgaJpZM4P3f0o.
Got it. Thanks.
On 19 Oct 2017, at 09:05, PatrickGreen123 notifications@github.com wrote:
Try this link:
https://drive.google.com/file/d/0B605tTCPt_5sT2JvRGswUW1STVE/view?usp=drivesdk
If that doesn't work then I'll try your suggestion
---- Markus Gesmann wrote ----
Hi Patrick, a made up triangle is exactly, what I am looking for. Unfortunately, it didn't seem to attached. Best, it you load into R and paste the output of
dput(triangle)
Thanks Markus
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHubhttps://github.com/mages/ChainLadder/issues/39#issuecomment-337373180, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AcwmAIrG-a-bVZSx3MUHNK_O-1lKm7r5ks5stRgsgaJpZM4P3f0o. — You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/mages/ChainLadder/issues/39#issuecomment-337831177, or mute the thread https://github.com/notifications/unsubscribe-auth/ABF-1YBwnztax_ID2tv7rRkq9IHQ9nT6ks5stwKvgaJpZM4P3f0o.
Updated package, version 0.2.5, on its way to CRAN. Thanks again.
The Mack method returns NaN for the the standard error if "sigma[i - 2]^2" is zero in the Mack.S.E function.