Cr\u00e9ons d\u2019abord un jeu de donn\u00e9es. On souhaite d\u00e9terminer la pr\u00e9cision et la r\u00e9p\u00e9tabilit\u00e9 d\u2019une analyse. Pour cela, la mesure est effectu\u00e9e par 2 techniciens diff\u00e9rents, sur des concentrations de produits diff\u00e9rents sur 3 jours diff\u00e9rents et avec 2 r\u00e9plicats.<\/p>\n
library(dplyr)\nset.seed(42)\ndata <- tibble(\n concentration = rep(c(10, 30, 50, 80), 3*2),\n technicien = rep(c(\"A\", \"B\"), each = 3*2*2),\n jour = rep(rep(1:3, each = 2*2), 2),\n replicat = rep(1:2, times = 2*3*2)) %>%\n mutate(mesure = ifelse(\n technicien == \"A\",\n 0.2 * concentration + rnorm(12, sd = 3),\n 0.2 * concentration + rnorm(12, sd = 2)))\ndata<\/code><\/pre>\n# A tibble: 24 x 5\n concentration technicien jour replicat mesure\n <dbl> <chr> <int> <int> <dbl>\n 1 10 A 1 1 7.69\n 2 30 A 1 2 4.71\n 3 50 A 1 1 9.23\n 5 10 A 2 1 3.38\n 6 30 A 2 2 4.08\n 9 10 A 3 1 5.11\n10 30 A 3 2 4.17\n# ... with 14 more rows<\/code><\/pre>\nAvec un effet al\u00e9atoire sur le jour, l\u2019anova peut se faire avec la fonction anovaVCA<\/code> du package {VCA}. Cette fonction, avec les param\u00e8tres par d\u00e9faut, r\u00e9alise exactement la m\u00eame Avona Type-I que la fonction PROC VARCOMP<\/code> dans SAS.<\/p>\nlibrary(VCA)\n# Result Variance Component Analysis\nres <- anovaVCA(mesure ~ (technicien)\/jour, \n Data = as.data.frame(data))\nres<\/code><\/pre>\nResult Variance Component Analysis:\n-----------------------------------\n\n Name DF SS MS \n1 total 17.345664 \n2 technicien 1 65.789545 65.789545\n3 technicien:jour 4 27.371049 6.842762 \n4 error 18 969.992448 53.888469\n\n VC %Total SD CV[%] \n1 58.800701 100 7.668162 84.168816\n2 4.912232 8.354036 2.216356 24.327607\n3 0* 0* 0* 0* \n4 53.888469 91.645964 7.340877 80.576405\n\nMean: 9.110454 (N = 24) \n\nExperimental Design: balanced | Method: ANOVA | * VC set to 0 | adapted MS used for total DF<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"Une anova avec mod\u00e8le mixte comme VARCOMP dans SAS Cr\u00e9ons d\u2019abord un jeu de donn\u00e9es. On souhaite d\u00e9terminer la pr\u00e9cision et la r\u00e9p\u00e9tabilit\u00e9 d\u2019une analyse. Pour cela, la mesure est effectu\u00e9e par 2 techniciens diff\u00e9rents, sur des concentrations de produits diff\u00e9rents sur 3 jours diff\u00e9rents et avec 2 r\u00e9plicats. library(dplyr) set.seed(42) data <- tibble( concentration = rep(c(10, 30, 50, 80), 3*2), technicien = rep(c(\u00ab\u00a0A\u00a0\u00bb, \u00ab\u00a0B\u00a0\u00bb), each = 3*2*2), jour = rep(rep(1:3, each = 2*2), 2), replicat = rep(1:2, times = 2*3*2)) %>% mutate(mesure = ifelse( technicien == \u00ab\u00a0A\u00a0\u00bb, 0.2 * concentration + rnorm(12, sd = 3), 0.2 * concentration + rnorm(12, sd = 2))) dataRead More →<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"content-type":"","rop_custom_images_group":[],"rop_custom_messages_group":[],"rop_publish_now":"initial","rop_publish_now_accounts":{"twitter_399453572_399453572":""},"rop_publish_now_history":[],"rop_publish_now_status":"pending","jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[30,31,14],"tags":[],"class_list":{"0":"entry","1":"post","2":"publish","3":"author-seb","4":"post-4560","6":"format-standard","7":"category-sas","8":"category-experimentation","9":"category-modelisation"},"acf":[],"jetpack_featured_media_url":"","jetpack_shortlink":"https:\/\/wp.me\/p9O7Sx-1by","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/posts\/4560","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/comments?post=4560"}],"version-history":[{"count":4,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/posts\/4560\/revisions"}],"predecessor-version":[{"id":4615,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/posts\/4560\/revisions\/4615"}],"wp:attachment":[{"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/media?parent=4560"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/categories?post=4560"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/thinkr.fr\/abcdr\/wp-json\/wp\/v2\/tags?post=4560"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}