AndyBunn
commented
1 year ago Alas though, people tend to prefer the “COFECHA” interseries correlation because the numbers are higher! In dplR that is the interseries.cor function. That stat is the correlation between a series and a master chronology.
From: Kevin Anchukaitis @.> Date: Thursday, March 30, 2023 at 4:10 PM To: OpenDendro/dplPy @.> Cc: Subscribed @.***> Subject: [OpenDendro/dplPy] Add calculation of running rbar for use in chronology statistics and variance stabilization (Issue #42)
rbar is typically calculated as 'the mean interseries correlation between all series from different trees' (Wigley et al. 1994). This can (and typically is) calculated over some common period that maximizes the number of years and series (e.g. the largest covering rectangle in a matrix of series (variables) and years (observations)). For use in deciding when there is sufficiently high, a running rbar can also be calculated for fixed windows of Y year overlapping by some number of years (e.g. 50 year windows lagged 25 years, or with 25 years of overlap; 100 year windows lagged 10 years at a time with 90 years of consecutive overlap, etc). dplR function is rwi.stats.running (https://github.com/cran/dplR/blob/master/R/rwi.stats.running.Rhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fcran%2FdplR%2Fblob%2Fmaster%2FR%2Frwi.stats.running.R&data=05%7C01%7Cbunna%40wwu.edu%7Cc40aa1e63ddf4a11588b08db3173f351%7Cdc46140ce26f43efb0ae00f257f478ff%7C0%7C0%7C638158146447461582%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=q%2FR4aDd94YBDnQ4MwFWqhBJrl36a8gn6Kg5BBXNQGvs%3D&reserved=0) Wigley1984.pdfhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FOpenDendro%2FdplPy%2Ffiles%2F11116825%2FWigley1984.pdf&data=05%7C01%7Cbunna%40wwu.edu%7Cc40aa1e63ddf4a11588b08db3173f351%7Cdc46140ce26f43efb0ae00f257f478ff%7C0%7C0%7C638158146447461582%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=KsAuwNRKrTu7nlaUH7f4vO3deFaaQzXvP7uF6xfMVhk%3D&reserved=0
— Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FOpenDendro%2FdplPy%2Fissues%2F42&data=05%7C01%7Cbunna%40wwu.edu%7Cc40aa1e63ddf4a11588b08db3173f351%7Cdc46140ce26f43efb0ae00f257f478ff%7C0%7C0%7C638158146447617825%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=PxmlVrSN9X4JJIGLZKxGR%2FgwhJFHFWwQzaWD1ekJwXA%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAC7UCXLHJANJWCGPXI5OW53W6YHGHANCNFSM6AAAAAAWN5S3WM&data=05%7C01%7Cbunna%40wwu.edu%7Cc40aa1e63ddf4a11588b08db3173f351%7Cdc46140ce26f43efb0ae00f257f478ff%7C0%7C0%7C638158146447617825%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=DHzHJZhfIBdF%2FgoTmZ6cSPl77ub8JWWnx4MbAJk%2F3sQ%3D&reserved=0. You are receiving this because you are subscribed to this thread.Message ID: @.***>
kanchukaitis
rbar is typically calculated as 'the mean interseries correlation between all series from different trees' (Wigley et al. 1994). This can (and typically is) calculated over some common period that maximizes the number of years and series (e.g. the largest covering rectangle in a matrix of series (variables) and years (observations)). For use in deciding when there is sufficiently high, a running rbar can also be calculated for fixed windows of Y year overlapping by some number of years (e.g. 50 year windows lagged 25 years, or with 25 years of overlap; 100 year windows lagged 10 years at a time with 90 years of consecutive overlap, etc). dplR function is rwi.stats.running (https://github.com/cran/dplR/blob/master/R/rwi.stats.running.R) Wigley1984.pdf