diff --git a/README.Rmd b/README.Rmd
index 614e3154673d367ce797ae0cd377e4550159bc0a..99a6ed7d9d9d9b3d75ee4def7c891a26766da03d 100644
--- a/README.Rmd
+++ b/README.Rmd
@@ -541,17 +541,27 @@ table(dat_select$ref_typoveg)
summary(dat_select)
# estimation du modèle
library(relaimpo)
+library(regclass)
select_mod = function(dat, dir = "both"){
mod_lm <- lm(norm_hmean_site ~ ., data =dat)
library(MASS)
modAIC = stepAIC(mod_lm, direction = dir)
df = broom::tidy(modAIC)
+ if(length(df$term)>2) {
+ vif = car::vif(modAIC)
+ vif_df = data.frame(term = names(vif), vif = vif)
+ relimp = relaimpo::calc.relimp(modAIC, rela = TRUE)
+ relimp_df = data.frame(term = relimp@namen[-1], relimp = relimp@lmg)
+ } else {
+ vif_df = data.frame(term = df$term, vif = NA)
+ relimp_df = data.frame(term = df$term, relimp = NA)
+ }
+
df$rsq = broom::glance(modAIC)$r.squared
- relimp = relaimpo::calc.relimp(modAIC, rela = TRUE)
- relimp_df = data.frame(term = relimp@namen[-1], relimp = relimp@lmg)
df = df %>% left_join(relimp_df)
- df$rsq2 = relimp@R2
+ df = df %>% left_join(vif_df)
+
return(df)
}
@@ -564,29 +574,113 @@ select_mod = function(dat, dir = "both"){
# dim(mod_season)
#
-mod_gperiod300 = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("LSD_safran"), contains("beginning"), contains("gdd300"), contains("precip300"), contains("Temp300") ))
+mod_gperiod300= dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, cumprecip_safran, contains("LSD_safran"), contains("speddgdd300_s"), contains("precip300"), contains("Temp300_T"), contains("ndayfrost300_T") ))
dim(mod_gperiod300)
+cor(mod_gperiod300[,-c(1:3)])
+
+mod_gperiod300_agromet = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, cumprecip_safran, contains("speedgdd300_s") ,contains("precip300"), contains("gdd300_agromet"), contains("Temp300_safran") ))
+dim(mod_gperiod300_agromet)
+cor(mod_gperiod300_agromet[,-c(1:3)])
mod_gperiod = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("LSD_safran"), contains("beginning"), contains("date_gdd300"), contains("period") , contains("cumgdd30") ))
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, cumprecip_safran, contains("LSD_safran"), contains("period"))) %>%
+ dplyr::select(-c("tgdd_period_safran"))
dim(mod_gperiod)
+cor(mod_gperiod[,-c(1:3)])
mod_previous30 = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("LSD_safran"), contains("beginning"), contains("date_gdd300") ,contains("previous30"), contains("cumgdd30") ))
-dim(mod_gperiod)
-
-mod_april = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("month4") ))
-dim(mod_april)
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("LSD_safran"), contains("previous30"), contains("cumgdd30_") ))
+dim(mod_previous30)
+cor(mod_previous30[,-c(1:3)])
+
+
+mod_april_min = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month4") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_april_min)
+cor(mod_april_min[,-c(1:3)])
+
+mod_april_max = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month4") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_april_max)
+cor(mod_april_max[,-c(1:3)])
+
+mod_april_soil = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month4") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_april_soil)
+cor(mod_april_soil[,-c(1:3)])
+
+
+
+mod_may_min = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month5") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_may_min)
+cor(mod_may_min[,-c(1:3)])
+
+mod_may_max = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month5") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_may_max)
+cor(mod_may_max[,-c(1:3)])
+
+mod_may_soil = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month5") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_may_soil)
+cor(mod_may_soil[,-c(1:3)])
+
+
+mod_june_min = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month6") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_june_min)
+cor(mod_june_min[,-c(1:3)])
+
+mod_june_max = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month6") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("soil")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_june_max)
+cor(mod_june_max[,-c(1:3)])
+
+mod_june_soil = dat_select %>%
+ dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumprecip_safran, contains("month6") )) %>%
+ dplyr::select(-contains("Tgdd_")) %>%
+ dplyr::select(-contains("Tmax")) %>%
+ dplyr::select(-contains("Tmin")) %>%
+ dplyr::select(-contains("runoff"))
+dim(mod_june_soil)
+cor(mod_june_soil[,-c(1:3)])
-mod_may = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("month5") ))
-dim(mod_may)
-mod_june = dat_select %>%
- dplyr::select(c(ref_typoveg, hmean,ref_site, minTemp_crosscut, maxTemp_crosscut, cumgdd_crosscut, cumprecip_safran, contains("month6") ))
-head(mod_june)
# mod_tmin = dat_select %>%
# dplyr::select(c(ref_typoveg, hmean,ref_site, contains("min") ))
@@ -604,7 +698,7 @@ head(mod_june)
# dplyr::select(c(ref_typoveg, hmean,ref_site, contains("tgdd"), contains("Tgdd"), contains("cumgdd") ))
# head(mod_gdd)
-mods_AIC_list = lapply(list(mod_gperiod, mod_gperiod300, mod_april, mod_may, mod_june), function(x){
+mods_AIC_list = lapply(list(mod_gperiod, mod_gperiod300_agromet, mod_gperiod300, mod_previous30, mod_april_min, mod_april_max, mod_may_min, mod_may_max, mod_june_min, mod_june_max), function(x){
x %>%
group_by(ref_site) %>%
mutate(site_mean = mean(hmean)) %>%
@@ -634,23 +728,21 @@ mods_AIC = rbind(
modsAIC_all = do.call(rbind, mods_AIC_list) %>% filter(!is.na(estimate)) %>% filter(p.value<0.2)
summary(modsAIC_all)
-
+modsAIC_all %>% dplyr::select(ref_typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
```
```{r , fig=TRUE, results='hide'}
-# require(ggplot2)
-#
-# mods_AIC_simplvars= mods_AIC %>%
-# mutate(term = if_else(grepl("LSD",term), "snow melt", term)) %>%
-# mutate(term = if_else(grepl("beginning",term), "snow melt", term)) %>%
-# mutate(term = if_else(grepl("date_gdd300",term), "speed gdd", term)) %>%
-# mutate(term = if_else(grepl("daygdd300",term), "speed gdd", term)) %>%
+# modsAIC_all= modsAIC_all %>% filter(p.value<0.1) %>%
+# mutate(term = if_else(grepl("LSD",term), " last snow day", term)) %>%
+# mutate(term = if_else(grepl("beginning",term), " last snow day", term)) %>%
+# mutate(term = if_else(grepl("date_gdd300",term), " speed gdd 300", term)) %>%
+# mutate(term = if_else(grepl("daygdd300",term), " speed gdd 300", term)) %>%
# mutate(term = if_else(grepl("cumgdd_period",term), "gdd growing start", term)) %>%
# mutate(term = if_else(grepl("precip_period",term), "precip growing start", term)) %>%
# mutate(term = if_else(grepl("min_period",term), "Tmin growing start", term)) %>%
# mutate(term = if_else(grepl("ndayfrost_period",term), "Tmin growing start", term)) %>%
-# mutate(term = if_else(grepl("gdd300",term), "Tmin growing start", term)) %>%
+# mutate(term = if_else(grepl("gdd300",term), "Tmin growing start", term)) %>%
# mutate(term = if_else(grepl("max_period",term), "Tmax growing start", term)) %>%
# mutate(term = if_else(grepl("tgdd_period",term), "gdd growing start", term)) %>%
# mutate(term = if_else(grepl("cumgdd30",term), "gdd growing start", term)) %>%
@@ -659,16 +751,20 @@ summary(modsAIC_all)
# mutate(term = if_else(grepl("minTemp300",term), "Tmin growing start", term)) %>%
# mutate(term = if_else(grepl("maxTemp300",term), "Tmax growing start", term)) %>%
#
+# mutate(term = if_else(grepl("ndayfrost300",term), "Tmin growing start", term)) %>%
+# mutate(term = if_else(grepl("previous30_runoff",term), " runoff 30d before", term)) %>%
+#
+#
#
# mutate(term = if_else(grepl("cumgdd",term), "gdd growing season", term)) %>%
# mutate(term = if_else(grepl("minTemp",term), "Tmin growing season", term)) %>%
# mutate(term = if_else(grepl("maxTemp",term), "Tmax growing season", term)) %>%
-#
-# mutate(term = if_else(grepl("ndayfrost",term), "Tmin growing season", term)) %>%
-#
+#
+# mutate(term = if_else(grepl("cumprecip",term), "precip growing season", term)) %>%
+#
# mutate(term = if_else(grepl("mm_month",term), "gdd growing start", term)) %>%
#
-# mutate(term = if_else(grepl("depth_month",term), "snow growing start", term)) %>%
+# mutate(term = if_else(grepl("depth_month",term), " snow quantity 4-6", term)) %>%
#
# mutate(term = if_else(grepl("max_month",term), "Tmax growing start", term)) %>%
#
@@ -676,23 +772,22 @@ summary(modsAIC_all)
#
# mutate(term = if_else(grepl("gdd_month",term), "gdd growing start", term)) %>%
#
-# mutate(term = if_else(grepl("precip_month",term), "precip growing start", term)) %>%
+# mutate(term = if_else(grepl("runoff_month",term), " runoff 4-6", term)) %>%
#
-# mutate(term = if_else(grepl("cumprecip",term), "precip growing season", term)) %>%
+# mutate(term = if_else(grepl("precip_month",term), "precip growing start", term)) %>%
#
# mutate(term = if_else(grepl("year",term), "year", term))
#
-# mods_AIC_simplvars$importance = (1-mods_AIC_simplvars$p.value)
-# imp_vars = mods_AIC_simplvars %>% filter(term != "(Intercept)") %>% group_by(ref_typoveg, term) %>% summarise(importance = sum(importance))
-#
+# imp_vars = modsAIC_all %>% filter(term != "(Intercept)") %>% group_by(ref_typoveg, term) %>% summarise(importance = sum(relimp))
#
-# ggplot(imp_vars, aes(x = "", y = importance, fill = term)) +
-# facet_wrap(~ref_typoveg) +
+# rsq_text = data.frame(ref_typoveg = list_rsq$ref_typoveg, label = paste0("rsq = ",round(list_rsq$rsq, 2)*100, "%"))
+#
+# ggplot(imp_vars %>% left_join(rsq_text), aes(x = "", y = importance, fill = term, label = label)) +
# geom_bar(stat = "identity") +
# coord_polar("y", start = 0) +
-# theme_void() +
-# scale_fill_brewer(palette = "Paired")
-#
+# theme_void() + scale_fill_brewer(palette="Paired") +
+# facet_wrap(~ref_typoveg) +
+# geom_text( aes(label= label), size = 3, x = 1, y = 0)
```
@@ -703,12 +798,76 @@ summary(modsAIC_all)
#modsAIC_all
final_mods = lapply(c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB"), function(x){
- select_terms = modsAIC_all %>% filter(ref_typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+ sel1 = modsAIC_all %>% filter(ref_typoveg == x) %>% filter(term!="(Intercept)")
+ if(max(sel1$vif , na.rm=TRUE)>2) {
+ select_terms = sel1 %>% filter(vif < max(vif, na.rm = TRUE)) %>% dplyr::select(term) %>% unique()
+ } else {
+ select_terms = modsAIC_all %>% filter(ref_typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+ }
+ dat_select %>% dplyr::select(c(select_terms$term, hmean,ref_site)) %>% mutate(site_mean = mean(hmean)) %>% mutate(norm_hmean_site = (hmean-site_mean)) %>% dplyr::select(-c(ref_site, site_mean, hmean)) %>% mutate_at(vars(-c(norm_hmean_site)), scale) %>% select_mod( dir ="both")
+ }) %>% bind_rows(.id="typoveg")
+final_mods$typoveg = as.factor(final_mods$typoveg)
+levels(final_mods$typoveg) = c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB")
+
+final_mods %>% dplyr::select(typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
+
+
+final_mods = lapply(c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB"), function(x){
+ sel1 = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)")
+ if(max(sel1$vif , na.rm=TRUE)>2) {
+ select_terms = sel1 %>% filter(vif < max(vif, na.rm = TRUE)) %>% dplyr::select(term) %>% unique()
+ } else {
+ select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+ }
dat_select %>% dplyr::select(c(select_terms$term, hmean,ref_site)) %>% mutate(site_mean = mean(hmean)) %>% mutate(norm_hmean_site = (hmean-site_mean)) %>% dplyr::select(-c(ref_site, site_mean, hmean)) %>% mutate_at(vars(-c(norm_hmean_site)), scale) %>% select_mod( dir ="both")
}) %>% bind_rows(.id="typoveg")
+final_mods$typoveg = as.factor(final_mods$typoveg)
+levels(final_mods$typoveg) = c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB")
+
+final_mods %>% dplyr::select(typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
+
+final_mods = lapply(c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB"), function(x){
+ if(max(final_mods %>% filter(typoveg == x) %>% dplyr::select(vif), na.rm=TRUE)>2) {
+ select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% filter(vif < max(vif, na.rm = TRUE)) %>% dplyr::select(term) %>% unique()
+ } else {
+ select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+ }
+ dat_select %>% dplyr::select(c(select_terms$term, hmean,ref_site)) %>% mutate(site_mean = mean(hmean)) %>% mutate(norm_hmean_site = (hmean-site_mean)) %>% dplyr::select(-c(ref_site, site_mean, hmean)) %>% mutate_at(vars(-c(norm_hmean_site)), scale) %>% select_mod( dir ="both")
+ }) %>% bind_rows(.id="typoveg")
+final_mods$typoveg = as.factor(final_mods$typoveg)
+levels(final_mods$typoveg) = c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB")
+
+final_mods %>% dplyr::select(typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
+final_mods = lapply(c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB"), function(x){
+ if(max(final_mods %>% filter(typoveg == x) %>% dplyr::select(vif), na.rm=TRUE)>2) {
+ select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% filter(vif < max(vif, na.rm = TRUE)) %>% dplyr::select(term) %>% unique()
+ } else {
+ select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+ }
+ dat_select %>% dplyr::select(c(select_terms$term, hmean,ref_site)) %>% mutate(site_mean = mean(hmean)) %>% mutate(norm_hmean_site = (hmean-site_mean)) %>% dplyr::select(-c(ref_site, site_mean, hmean)) %>% mutate_at(vars(-c(norm_hmean_site)), scale) %>% select_mod( dir ="both")
+ }) %>% bind_rows(.id="typoveg")
final_mods$typoveg = as.factor(final_mods$typoveg)
levels(final_mods$typoveg) = c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB")
+
+final_mods %>% dplyr::select(typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
+
+#
+# final_mods = lapply(c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB"), function(x){
+# if(max(final_mods %>% filter(typoveg == x) %>% dplyr::select(vif), na.rm=TRUE)>2) {
+# select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% filter(vif < max(vif)) %>% dplyr::select(term) %>% unique()
+# } else {
+# select_terms = final_mods %>% filter(typoveg == x) %>% filter(term!="(Intercept)") %>% dplyr::select(term) %>% unique()
+# }
+# dat_select %>% dplyr::select(c(select_terms$term, hmean,ref_site)) %>% mutate(site_mean = mean(hmean)) %>% mutate(norm_hmean_site = (hmean-site_mean)) %>% dplyr::select(-c(ref_site, site_mean, hmean)) %>% mutate_at(vars(-c(norm_hmean_site)), scale) %>% select_mod( dir ="both")
+# }) %>% bind_rows(.id="typoveg")
+# final_mods$typoveg = as.factor(final_mods$typoveg)
+# levels(final_mods$typoveg) = c("ALP", "ECOR","NIV", "ENHE", "NAR", "BOMB", "PROD","QUE", "SUB")
+#
+# final_mods %>% dplyr::select(typoveg, term, vif) %>% filter(vif>2) %>% arrange(desc(vif))
+#
+
+
```
```{r}
@@ -719,7 +878,7 @@ rsq = final_mods %>% dplyr::select(typoveg, rsq) %>% group_by(typoveg) %>% uniqu
#### Par groupe de variables par milieu
-```{r , fig=TRUE, results='hide'}
+```{r final , fig=TRUE, results='hide'}
require(ggplot2)
summary(final_mods)
diff --git a/README.md b/README.md
index f03e827214c1790114c03f2da6e2966b5f69e63d..94412107a366f4acba30a16a6452c7abdd4017ec 100644
--- a/README.md
+++ b/README.md
@@ -72,7 +72,7 @@ Il s'agit de :
## 8 ECRGRA03 0.667 QUE
## 9 ECRGRA04 0.630 ALP
## 10 ECRGRA05 0.899 ALP
-## # … with 45 more rows
+## # ℹ 45 more rows
```
```
@@ -295,7 +295,7 @@ Les mesures réalisées trop tôt ou trop tard sont pas exploitables. Mieux vaut
## 8 BELRIV02 3 QUE
## 9 BELRIV03 3 SUB
## 10 ECRSUR03 3 SUB
-## # … with 11 more rows
+## # ℹ 11 more rows
```
A réfléchir comment gérer les contraintes...
@@ -312,8 +312,8 @@ A réfléchir comment gérer les contraintes...
## Number of trees: 100
## No. of variables tried at each split: 16
##
-## Mean of squared residuals: 10.85859
-## % Var explained: 83.54
+## Mean of squared residuals: 11.41521
+## % Var explained: 82.69
```
<!-- -->
@@ -361,7 +361,7 @@ Analyse par type de facteurs
#### Par groupe de variables par milieu
-<!-- -->
+<!-- -->
On retrouve une importance des températures (mini, maxi, ...) pour la plupart des milieux.
- Tmax pour ECOR, NIV ENHE et PROD et SUB
@@ -444,12 +444,12 @@ Les Nardaies
## Effets spatiaux purs: typo de végétation et nos variables
-<!-- -->
+<!-- -->
On a surtout un axe qui ressort et est en lien avec les variables de température. De façon intéressante, le deuxième axe fait ressortir les minimales/ gel avec un large degré d'indépendance donc.
-<!-- -->
+<!-- -->
Les types de végétation se répartissent assez bien le long de cet axe principal, avec une position similaire pour PROD, QUE, NAR, et ECOR, qui se différencient plus ou moins sur leur tolérance au gel.
@@ -457,14 +457,14 @@ Les PROD sont les plus hétérogènes sur l'axe 1.
Les PROD/NAR et QUE et ECOR se distinguent par l'axe 2 (gel/ temp mini)
Répartition des sites:
-<!-- -->
+<!-- -->
Discimination des types de végétation:
Qu'est ce qui disciminent le plus les types de végétation?
-<!-- --><!-- -->
+<!-- --><!-- -->
Encore une fois, on retrouve le fait que QUE/PROD et NAR sont peu discriminées par les variables env., surtout PROD englobe NAR et QUE.
@@ -488,12 +488,12 @@ Encore une fois, on retrouve le fait que QUE/PROD et NAR sont peu discriminées
## Effets années
-<!-- -->
+<!-- -->
En prenant en compte les effets temporels, les axes restent du même genre mais on a moins de corrélations.
-<!-- -->
+<!-- -->
On retrouve ici des années typiques :
diff --git a/README_files/figure-html/generalModel-1.png b/README_files/figure-html/generalModel-1.png
index 4073b5d83f34ca14dc42ee89a306752b778c1e77..4b955311b2c1ca7171bf3ab5a6ea9230d4290f4f 100644
Binary files a/README_files/figure-html/generalModel-1.png and b/README_files/figure-html/generalModel-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-19-1.png b/README_files/figure-html/unnamed-chunk-19-1.png
index 41474c8eaa7178bad0c17e673c195efc39114552..37ce40f3a5b85386ffbeab11dd492cfa03ef790a 100644
Binary files a/README_files/figure-html/unnamed-chunk-19-1.png and b/README_files/figure-html/unnamed-chunk-19-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-22-1.png b/README_files/figure-html/unnamed-chunk-22-1.png
index 22b048fa0ad313231f3211fa2df1a88cba3f8cd5..8642f441856f3c0d6f1ca328a6545bd70f4ddb04 100644
Binary files a/README_files/figure-html/unnamed-chunk-22-1.png and b/README_files/figure-html/unnamed-chunk-22-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-24-1.png b/README_files/figure-html/unnamed-chunk-24-1.png
index 054e128eb2671cd6bd1fb305785e79559546b39b..31126e23c878f0ced71a90f0b8a0d9ee44619926 100644
Binary files a/README_files/figure-html/unnamed-chunk-24-1.png and b/README_files/figure-html/unnamed-chunk-24-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-25-1.png b/README_files/figure-html/unnamed-chunk-25-1.png
index 076fc0283acf7b5848f57ea3d5a3e4eb11673bb2..819af8b13592fc6edd74b380937caa566eedfcd6 100644
Binary files a/README_files/figure-html/unnamed-chunk-25-1.png and b/README_files/figure-html/unnamed-chunk-25-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-26-1.png b/README_files/figure-html/unnamed-chunk-26-1.png
index 819af8b13592fc6edd74b380937caa566eedfcd6..defa0e7f66ed9de104e5a7c8e4ca2217a879c20b 100644
Binary files a/README_files/figure-html/unnamed-chunk-26-1.png and b/README_files/figure-html/unnamed-chunk-26-1.png differ
diff --git a/README_files/figure-html/unnamed-chunk-27-1.png b/README_files/figure-html/unnamed-chunk-27-1.png
index 3a7a53cda3d216dc1da675e3d094655665f464de..c0bae96240bfc42bf263e3b6a22ca8e43550a974 100644
Binary files a/README_files/figure-html/unnamed-chunk-27-1.png and b/README_files/figure-html/unnamed-chunk-27-1.png differ