Morts Collisions
2025-08-01
1 Importations / nettoyage
library(ggfortify)
library(lmtest)
library(dplyr)
library(ggplot2)wd <- set_wd()Total_et_Diro <- readr::read_csv(paste0(wd$data,
"derived/Total_et_Diro.csv"),
locale = readr::locale(encoding = "UTF-8"))## New names:
## Rows: 966902 Columns: 16
## ── Column specification
## ───────────────────────────────────────────────────────────────────── Delimiter: "," chr
## (5): Code_10km, paysage_ID, paysage_nom, famille_paysage, etat_biologique dbl (10): ...1,
## X_10km, Y_10km, cd_nom, cd_nom_grp_must, Indice_Diversite, Densite_Cultu... date (1): date
## ℹ Use `spec()` to retrieve the full column specification for this data. ℹ Specify the column
## types or set `show_col_types = FALSE` to quiet this message.
## • `` -> `...1`Total_et_Diro <- Total_et_Diro %>%
dplyr::select(-`...1`)2 Mortalit?s routi?res - Toutes donn?es
Donnees <- Total_et_Diro %>%
dplyr::filter(etat_biologique == "Trouvé mort : impact routier")2.1 Lapin
2.1.1 Lapin - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 61714,
espece_benchmark = 60981)liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -465.507regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-7
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-8
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Ind_Diversite + Dnst_Cultures + Dist_Ecotone + Dist_Littoral + prop_tmoins1 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Ind_Diversite +
## Dnst_Cultures + Dist_Ecotone + Dist_Littoral + prop_tmoins1 +
## year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.288e+04 2.712e+03 4.750 2.26e-06 ***
## year -1.272e+01 2.688e+00 -4.733 2.46e-06 ***
## X_10km -1.160e-01 1.454e-02 -7.979 3.25e-15 ***
## Ind_Diversite 4.460e-01 9.458e-02 4.716 2.67e-06 ***
## Dnst_Cultures 5.897e-01 1.573e-01 3.750 0.000185 ***
## Dist_Ecotone 5.285e-03 8.419e-04 6.277 4.70e-10 ***
## Dist_Littoral 4.456e-03 7.821e-04 5.698 1.50e-08 ***
## prop_tmoins1 1.716e-01 2.377e-02 7.219 8.94e-13 ***
## year2 3.141e-03 6.661e-04 4.715 2.68e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1173963)
##
## Null deviance: 227.64 on 1297 degrees of freedom
## Residual deviance: 151.32 on 1289 degrees of freedom
## AIC: 913.96
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-10
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 6.9345, df = 1, p-value = 0.008455##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 31.809, df = 8, p-value = 0.0001008## year X_10km Ind_Diversite Dnst_Cultures Dist_Ecotone Dist_Littoral
## 1.201113e+06 1.941869e+00 2.948496e+00 2.462200e+00 1.103397e+00 2.281362e+00
## prop_tmoins1 year2
## 1.169420e+00 1.201139e+062.1.2 Lapin - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 61714,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -719.632regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-12
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-13
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dist_Littoral + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dist_Littoral + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.1338852 0.0389847 3.434 0.000606 ***
## year2011 0.0092822 0.0475941 0.195 0.845390
## year2012 -0.0541655 0.0476910 -1.136 0.256194
## year2013 -0.1190101 0.0479806 -2.480 0.013206 *
## year2014 0.1500649 0.0472520 3.176 0.001517 **
## year2015 0.1849145 0.0490890 3.767 0.000170 ***
## year2016 0.0010400 0.0497601 0.021 0.983327
## year2017 0.0184133 0.0499463 0.369 0.712419
## year2018 0.0443687 0.0500759 0.886 0.375709
## year2019 0.0727563 0.0511132 1.423 0.154766
## year2020 0.0976862 0.0514856 1.897 0.057926 .
## year2021 0.0279876 0.0525967 0.532 0.594704
## year2022 -0.0022390 0.0515566 -0.043 0.965365
## year2023 0.0228052 0.0526237 0.433 0.664798
## year2024 -0.0644338 0.0523166 -1.232 0.218239
## year2025 -0.3346137 0.0687775 -4.865 1.23e-06 ***
## Dist_Littoral 0.0017440 0.0004679 3.727 0.000199 ***
## prop_tmoins1 0.3536948 0.0234982 15.052 < 2e-16 ***
## prop_tmoins2 0.2329948 0.0250089 9.316 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.129139)
##
## Null deviance: 383.87 on 2014 degrees of freedom
## Residual deviance: 257.76 on 1996 degrees of freedom
## AIC: 1614.8
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-15
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 3.1589, df = 1, p-value = 0.07551##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 54.598, df = 18, p-value = 1.48e-05## GVIF Df GVIF^(1/(2*Df))
## year -0.1317796 15 NaN
## Dist_Littoral 0.1685464 1 0.4105440
## prop_tmoins1 0.1462464 1 0.3824218
## prop_tmoins2 0.1545941 1 0.39318452.1.3 Lapin - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 61714,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -446.8622regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-17
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-18
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dnst_Cultures + Dist_Ecotone + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dnst_Cultures +
## Dist_Ecotone + prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.098e+04 2.371e+03 8.849 < 2e-16 ***
## year -2.075e+01 2.350e+00 -8.833 < 2e-16 ***
## X_10km -4.135e-02 1.024e-02 -4.039 5.63e-05 ***
## Dnst_Cultures 2.390e-01 9.803e-02 2.438 0.01488 *
## Dist_Ecotone 5.928e-03 7.602e-04 7.798 1.18e-14 ***
## prop_tmoins1 1.501e-01 2.427e-02 6.183 8.12e-10 ***
## prop_tmoins2 8.563e-02 2.524e-02 3.393 0.00071 ***
## year2 5.133e-03 5.823e-04 8.816 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1074528)
##
## Null deviance: 222.83 on 1482 degrees of freedom
## Residual deviance: 158.49 on 1475 degrees of freedom
## AIC: 910.42
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-20
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 12.827, df = 1, p-value = 0.0003416##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 42.784, df = 7, p-value = 3.672e-07## year X_10km Dnst_Cultures Dist_Ecotone prop_tmoins1 prop_tmoins2
## 1.150401e+06 1.189791e+00 1.198093e+00 1.075757e+00 1.165040e+00 1.190092e+00
## year2
## 1.150371e+062.1.4 Lapin - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 61714,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -385.5912regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-22
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-23
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Ind_Diversite + Dist_Littoral + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Ind_Diversite + Dist_Littoral +
## prop_tmoins1 + prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 35.6849378 2.7279335 13.081 < 2e-16 ***
## year -0.0177765 0.0013522 -13.146 < 2e-16 ***
## Ind_Diversite 0.1371118 0.0403605 3.397 0.000692 ***
## Dist_Littoral 0.0016922 0.0003863 4.381 1.23e-05 ***
## prop_tmoins1 0.1887340 0.0197779 9.543 < 2e-16 ***
## prop_tmoins2 0.1964484 0.0201291 9.759 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.08111557)
##
## Null deviance: 237.47 on 2461 degrees of freedom
## Residual deviance: 199.22 on 2456 degrees of freedom
## AIC: 810.6
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-25
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 4.6189, df = 1, p-value = 0.03162##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 254.33, df = 5, p-value < 2.2e-16## year Ind_Diversite Dist_Littoral prop_tmoins1 prop_tmoins2
## 1.011351 1.401660 1.420254 1.109175 1.1099882.2 Herisson
2.2.1 Herisson - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 60015,
espece_benchmark = 60981)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -580.384regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-27
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-28
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dnst_Cultures + Dist_Ecotone + Dist_Littoral + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dnst_Cultures +
## Dist_Ecotone + Dist_Littoral + prop_tmoins1 + prop_tmoins2 +
## year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.864e+04 1.313e+03 14.197 < 2e-16 ***
## year -1.846e+01 1.301e+00 -14.184 < 2e-16 ***
## X_10km -3.409e-02 7.804e-03 -4.368 1.30e-05 ***
## Dnst_Cultures -3.038e-01 6.001e-02 -5.061 4.44e-07 ***
## Dist_Ecotone 1.833e-03 3.598e-04 5.095 3.72e-07 ***
## Dist_Littoral 1.709e-03 3.937e-04 4.340 1.48e-05 ***
## prop_tmoins1 2.154e-01 1.552e-02 13.882 < 2e-16 ***
## prop_tmoins2 6.566e-02 1.510e-02 4.350 1.41e-05 ***
## year2 4.569e-03 3.225e-04 14.171 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.07432806)
##
## Null deviance: 257.42 on 2738 degrees of freedom
## Residual deviance: 202.92 on 2730 degrees of freedom
## AIC: 664.54
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-30
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 25.174, df = 1, p-value = 5.238e-07##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 269.85, df = 8, p-value < 2.2e-16## year X_10km Dnst_Cultures Dist_Ecotone Dist_Littoral prop_tmoins1
## 1.216228e+06 2.011004e+00 1.297551e+00 1.143132e+00 1.949767e+00 1.374611e+00
## prop_tmoins2 year2
## 1.624175e+00 1.215901e+062.2.2 Herisson - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60015,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -209.4285regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-32
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-33
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Y_10km + Dist_Ecotone + prop_tmoins1 + prop_tmoins2 )summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Y_10km + Dist_Ecotone +
## prop_tmoins1 + prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.5266990 0.9577796 4.726 2.39e-06 ***
## year2011 0.0116033 0.0391522 0.296 0.766973
## year2012 0.0061287 0.0395622 0.155 0.876901
## year2013 -0.1069756 0.0409692 -2.611 0.009070 **
## year2014 -0.0192134 0.0404588 -0.475 0.634902
## year2015 -0.0143612 0.0424891 -0.338 0.735389
## year2016 -0.0660909 0.0430372 -1.536 0.124726
## year2017 0.0098551 0.0423201 0.233 0.815878
## year2018 0.0814820 0.0415810 1.960 0.050136 .
## year2019 0.1545651 0.0417786 3.700 0.000220 ***
## year2020 0.1143825 0.0427767 2.674 0.007537 **
## year2021 0.1277271 0.0431242 2.962 0.003082 **
## year2022 0.1144934 0.0429157 2.668 0.007675 **
## year2023 0.0958650 0.0435086 2.203 0.027646 *
## year2024 0.0719818 0.0430337 1.673 0.094495 .
## year2025 0.0798604 0.0480563 1.662 0.096656 .
## X_10km 0.0282742 0.0072479 3.901 9.79e-05 ***
## Y_10km -0.0819416 0.0200027 -4.097 4.31e-05 ***
## Dist_Ecotone 0.0014384 0.0004189 3.434 0.000604 ***
## prop_tmoins1 0.1009301 0.0181757 5.553 3.05e-08 ***
## prop_tmoins2 0.0599650 0.0182682 3.282 0.001041 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.122626)
##
## Null deviance: 407.63 on 3003 degrees of freedom
## Residual deviance: 365.79 on 2983 degrees of freedom
## AIC: 2243.7
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-35
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 4.7931, df = 1, p-value = 0.02857##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 153.12, df = 20, p-value < 2.2e-16## GVIF Df GVIF^(1/(2*Df))
## year -1.081988 15 NaN
## X_10km 2.014977 1 1.419499
## Y_10km 1.860013 1 1.363823
## Dist_Ecotone 1.969876 1 1.403523
## prop_tmoins1 1.321276 1 1.149468
## prop_tmoins2 1.215757 1 1.1026142.2.3 Herisson - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60015,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -859.524regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-37
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-38
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dnst_Cultures + Dist_Ecotone + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dnst_Cultures + Dist_Ecotone +
## prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.673e+04 1.395e+03 19.163 < 2e-16 ***
## year -2.648e+01 1.383e+00 -19.149 < 2e-16 ***
## Dnst_Cultures -2.515e-01 5.890e-02 -4.270 2.02e-05 ***
## Dist_Ecotone 2.605e-03 3.802e-04 6.851 8.94e-12 ***
## prop_tmoins1 2.873e-01 1.622e-02 17.710 < 2e-16 ***
## prop_tmoins2 7.908e-02 1.609e-02 4.914 9.41e-07 ***
## year2 6.556e-03 3.426e-04 19.135 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.0929708)
##
## Null deviance: 363.88 on 2843 degrees of freedom
## Residual deviance: 263.76 on 2837 degrees of freedom
## AIC: 1324.1
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-40
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 58.427, df = 1, p-value = 2.11e-14##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 219.7, df = 6, p-value < 2.2e-16## year Dnst_Cultures Dist_Ecotone prop_tmoins1 prop_tmoins2 year2
## 1.122100e+06 1.028890e+00 1.028776e+00 1.350727e+00 1.513219e+00 1.121935e+062.2.4 Herisson - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60015,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -323.177regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-42
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-43
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dnst_Cultures + Dist_Ecotone + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dnst_Cultures +
## Dist_Ecotone + prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.283e+04 1.507e+03 8.512 < 2e-16 ***
## year -1.270e+01 1.494e+00 -8.500 < 2e-16 ***
## X_10km 5.363e-02 7.678e-03 6.985 3.44e-12 ***
## Dnst_Cultures -5.726e-01 7.256e-02 -7.891 4.05e-15 ***
## Dist_Ecotone 2.475e-03 4.214e-04 5.873 4.69e-09 ***
## prop_tmoins1 1.533e-01 1.681e-02 9.119 < 2e-16 ***
## prop_tmoins2 5.658e-02 1.709e-02 3.310 0.000944 ***
## year2 3.142e-03 3.702e-04 8.488 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1349149)
##
## Null deviance: 498.88 on 3294 degrees of freedom
## Residual deviance: 443.47 on 3287 degrees of freedom
## AIC: 2760.5
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-45
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 14.441, df = 1, p-value = 0.0001446##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 27.444, df = 7, p-value = 0.0002772## year X_10km Dnst_Cultures Dist_Ecotone prop_tmoins1 prop_tmoins2
## 1.080092e+06 1.290904e+00 1.214899e+00 1.138351e+00 1.110128e+00 1.189957e+00
## year2
## 1.079997e+062.3 Putois
2.3.1 Putois - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 60731,
espece_benchmark = 60981)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -252.7188regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-47
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-48
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dist_Ecotone + prop_tmoins1 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dist_Ecotone +
## prop_tmoins1 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.065e+04 3.393e+03 9.035 < 2e-16 ***
## year -3.035e+01 3.362e+00 -9.026 < 2e-16 ***
## X_10km -7.974e-02 1.337e-02 -5.966 3.32e-09 ***
## Dist_Ecotone 1.693e-03 1.019e-03 1.661 0.097 .
## prop_tmoins1 2.228e-01 3.075e-02 7.246 8.29e-13 ***
## year2 7.511e-03 8.330e-04 9.017 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1445097)
##
## Null deviance: 201.10 on 1059 degrees of freedom
## Residual deviance: 152.31 on 1054 degrees of freedom
## AIC: 965.66
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-50
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 3.8472, df = 1, p-value = 0.04983##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 29.957, df = 5, p-value = 1.504e-05## year X_10km Dist_Ecotone prop_tmoins1 year2
## 1.145774e+06 1.098791e+00 1.032899e+00 1.052894e+00 1.145752e+062.3.2 Putois - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60731,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -60.71499regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-52
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-53
reg <- glm(data = bdd,
formula = proportion_interet ~ year + prop_tmoins1)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + prop_tmoins1, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.100000 0.037074 2.697 0.00706 **
## year2011 -0.021091 0.046481 -0.454 0.65007
## year2012 -0.014730 0.046348 -0.318 0.75066
## year2013 -0.044214 0.046833 -0.944 0.34527
## year2014 0.078493 0.046497 1.688 0.09157 .
## year2015 0.081432 0.050345 1.617 0.10596
## year2016 0.119862 0.048878 2.452 0.01430 *
## year2017 0.003373 0.050618 0.067 0.94688
## year2018 0.088713 0.049791 1.782 0.07497 .
## year2019 0.100960 0.052074 1.939 0.05270 .
## year2020 0.172509 0.053208 3.242 0.00121 **
## year2021 0.173524 0.052765 3.289 0.00103 **
## year2022 0.292060 0.049710 5.875 5.08e-09 ***
## year2023 0.178150 0.052209 3.412 0.00066 ***
## year2024 0.066053 0.051030 1.294 0.19570
## year2025 0.016859 0.062925 0.268 0.78880
## prop_tmoins1 0.079268 0.027340 2.899 0.00379 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.115454)
##
## Null deviance: 209.59 on 1690 degrees of freedom
## Residual deviance: 193.27 on 1674 degrees of freedom
## AIC: 1167.1
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-55
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 1.1152, df = 1, p-value = 0.291##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 111.97, df = 16, p-value < 2.2e-16## GVIF Df GVIF^(1/(2*Df))
## year 0.533688 15 0.9792861
## prop_tmoins1 0.533688 1 0.73053952.3.3 Putois - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60731,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -243.655regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-57
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-58
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral +
## prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.443e+04 2.619e+03 9.329 < 2e-16 ***
## year -2.419e+01 2.595e+00 -9.322 < 2e-16 ***
## Dist_Ecotone 2.440e-03 8.034e-04 3.037 0.00243 **
## Dist_Littoral -2.147e-03 5.161e-04 -4.159 3.38e-05 ***
## prop_tmoins1 2.554e-01 2.910e-02 8.774 < 2e-16 ***
## prop_tmoins2 1.006e-01 3.320e-02 3.032 0.00247 **
## year2 5.989e-03 6.430e-04 9.315 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1185592)
##
## Null deviance: 205.78 on 1416 degrees of freedom
## Residual deviance: 167.17 on 1410 degrees of freedom
## AIC: 1008.7
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-60
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 10.113, df = 1, p-value = 0.001472##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 88.521, df = 6, p-value < 2.2e-16## year Dist_Ecotone Dist_Littoral prop_tmoins1 prop_tmoins2 year2
## 1.115341e+06 1.009039e+00 1.038932e+00 1.118044e+00 1.111462e+00 1.115320e+062.3.4 Putois - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60731,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -8.083715regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-62
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-63
reg <- glm(data = bdd,
formula = proportion_interet ~ year + prop_tmoins1 )summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + prop_tmoins1, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 11.027184 2.266871 4.864 1.22e-06 ***
## year -0.005422 0.001123 -4.828 1.47e-06 ***
## prop_tmoins1 0.061226 0.021283 2.877 0.00405 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.05335957)
##
## Null deviance: 128.84 on 2385 degrees of freedom
## Residual deviance: 127.16 on 2383 degrees of freedom
## AIC: -216.48
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-65
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 8.3756, df = 1, p-value = 0.003803##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 22.995, df = 2, p-value = 1.016e-05## year prop_tmoins1
## 1 12.4 Hermine
2.4.1 Hermine - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 60686,
espece_benchmark = 60981)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -67.65313regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-67
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-68
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dist_Littoral + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dist_Littoral +
## year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.179e+04 2.548e+03 4.626 4.35e-06 ***
## year -1.166e+01 2.524e+00 -4.619 4.49e-06 ***
## X_10km -4.401e-02 9.883e-03 -4.453 9.68e-06 ***
## Dist_Littoral 8.823e-04 4.735e-04 1.863 0.0628 .
## year2 2.883e-03 6.251e-04 4.613 4.63e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.03627921)
##
## Null deviance: 32.871 on 803 degrees of freedom
## Residual deviance: 28.987 on 799 degrees of freedom
## AIC: -377.84
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-70
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 0.63783, df = 1, p-value = 0.4245##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 87.413, df = 4, p-value < 2.2e-16## year X_10km Dist_Littoral year2
## 1.515306e+06 1.742716e+00 1.736049e+00 1.515317e+062.4.2 Hermine - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60686,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -3.300864regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-72
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-73
reg <- glm(data = bdd,
formula = proportion_interet ~ year + prop_tmoins1)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + prop_tmoins1, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.021097 0.018638 1.132 0.25784
## year2011 -0.020977 0.023313 -0.900 0.36839
## year2012 -0.017705 0.023222 -0.762 0.44593
## year2013 -0.019022 0.023341 -0.815 0.41522
## year2014 0.051367 0.023372 2.198 0.02812 *
## year2015 0.039185 0.025271 1.551 0.12121
## year2016 0.022565 0.024955 0.904 0.36602
## year2017 0.060731 0.025191 2.411 0.01604 *
## year2018 0.014796 0.025235 0.586 0.55773
## year2019 0.027987 0.026642 1.050 0.29367
## year2020 0.031437 0.027869 1.128 0.25948
## year2021 0.001926 0.027483 0.070 0.94413
## year2022 0.014168 0.026556 0.534 0.59376
## year2023 -0.009695 0.027425 -0.354 0.72374
## year2024 -0.024307 0.025984 -0.935 0.34971
## year2025 -0.021097 0.032147 -0.656 0.51175
## prop_tmoins1 0.089876 0.029351 3.062 0.00224 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.02744253)
##
## Null deviance: 42.267 on 1502 degrees of freedom
## Residual deviance: 40.780 on 1486 degrees of freedom
## AIC: -1120
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-75
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 0.27612, df = 1, p-value = 0.5993##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 48.619, df = 16, p-value = 3.795e-05## GVIF Df GVIF^(1/(2*Df))
## year 0.5031592 15 0.9773651
## prop_tmoins1 0.5031592 1 0.70933722.4.3 Hermine - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60686,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -17.70485regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-77
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-78
reg <- glm(data = bdd,
formula = proportion_interet ~ year + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.276e+03 1.220e+03 2.684 0.00737 **
## year -3.240e+00 1.209e+00 -2.679 0.00748 **
## year2 8.010e-04 2.995e-04 2.674 0.00760 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.01692391)
##
## Null deviance: 20.706 on 1186 degrees of freedom
## Residual deviance: 20.038 on 1184 degrees of freedom
## AIC: -1468.2
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-80
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 0.02438, df = 1, p-value = 0.8759##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 32.48, df = 2, p-value = 8.853e-08## year year2
## 1226299 12262992.4.4 Hermine - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60686,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] 1.606498regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-82
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-83
reg <- glm(data = bdd,
formula = proportion_interet ~ year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.403e+00 3.738e-01 3.753 0.000179 ***
## year2 -3.420e-07 9.176e-08 -3.728 0.000198 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.005487565)
##
## Null deviance: 12.621 on 2287 degrees of freedom
## Residual deviance: 12.545 on 2286 degrees of freedom
## AIC: -5412.6
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-85
bgtest(reg); bptest(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 0.97838, df = 1, p-value = 0.3226##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 8.3927, df = 1, p-value = 0.0037672.5 Belette
2.5.1 Belette - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 60716,
espece_benchmark = 60981)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -187.7034regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-87
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-88
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dnst_Cultures + Dist_Ecotone + Dist_Eau + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dnst_Cultures +
## Dist_Ecotone + Dist_Eau + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.216e+04 3.412e+03 6.495 1.32e-10 ***
## year -2.192e+01 3.381e+00 -6.484 1.41e-10 ***
## X_10km -3.995e-02 1.456e-02 -2.745 0.006163 **
## Dnst_Cultures -6.779e-01 1.436e-01 -4.722 2.68e-06 ***
## Dist_Ecotone 4.189e-03 1.103e-03 3.797 0.000156 ***
## Dist_Eau 1.002e-03 2.783e-04 3.601 0.000333 ***
## year2 5.422e-03 8.376e-04 6.474 1.51e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1316313)
##
## Null deviance: 163.71 on 985 degrees of freedom
## Residual deviance: 128.87 on 979 degrees of freedom
## AIC: 807.76
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-90
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 2.1107, df = 1, p-value = 0.1463##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 37.005, df = 6, p-value = 1.757e-06## year X_10km Dnst_Cultures Dist_Ecotone Dist_Eau year2
## 1.120692e+06 1.285607e+00 1.388575e+00 1.133976e+00 1.145289e+00 1.120678e+062.5.2 Belette - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60716,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -87.84112regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-92
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-93
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + prop_tmoins1 + prop_tmoins2 )summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.173502 0.045371 3.824 0.000136 ***
## year2011 -0.020876 0.044504 -0.469 0.639070
## year2012 -0.034549 0.044670 -0.773 0.439378
## year2013 -0.090631 0.044997 -2.014 0.044155 *
## year2014 0.064516 0.044572 1.447 0.147960
## year2015 0.106576 0.047294 2.253 0.024361 *
## year2016 0.114972 0.046694 2.462 0.013908 *
## year2017 0.048587 0.048164 1.009 0.313224
## year2018 0.088171 0.047693 1.849 0.064675 .
## year2019 0.094879 0.049991 1.898 0.057877 .
## year2020 0.160823 0.050801 3.166 0.001575 **
## year2021 0.105226 0.050833 2.070 0.038605 *
## year2022 -0.021229 0.050965 -0.417 0.677065
## year2023 0.053067 0.051164 1.037 0.299795
## year2024 -0.006088 0.049549 -0.123 0.902231
## year2025 0.026291 0.059542 0.442 0.658867
## X_10km 0.027174 0.008877 3.061 0.002240 **
## prop_tmoins1 0.132587 0.027333 4.851 1.34e-06 ***
## prop_tmoins2 0.147530 0.030794 4.791 1.81e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1052551)
##
## Null deviance: 192.41 on 1670 degrees of freedom
## Residual deviance: 173.88 on 1652 degrees of freedom
## AIC: 1000.9
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-95
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 0.49732, df = 1, p-value = 0.4807##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 114.4, df = 18, p-value = 4.734e-16## GVIF Df GVIF^(1/(2*Df))
## year -0.004020468 15 NaN
## X_10km 0.330103001 1 0.5745459
## prop_tmoins1 0.338503820 1 0.5818108
## prop_tmoins2 0.334950227 1 0.57874882.5.3 Belette - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60716,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -165.6966regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-97
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-98
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral + prop_tmoins1 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral +
## prop_tmoins1 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.093e+04 2.356e+03 8.883 < 2e-16 ***
## year -2.072e+01 2.335e+00 -8.875 < 2e-16 ***
## Dist_Ecotone 2.674e-03 7.491e-04 3.569 0.000371 ***
## Dist_Littoral -2.448e-03 4.550e-04 -5.380 8.82e-08 ***
## prop_tmoins1 1.066e-01 3.046e-02 3.498 0.000484 ***
## year2 5.129e-03 5.784e-04 8.867 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.08760732)
##
## Null deviance: 134.15 on 1311 degrees of freedom
## Residual deviance: 114.42 on 1306 degrees of freedom
## AIC: 536.7
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-100
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 7.0847, df = 1, p-value = 0.007775##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 130.12, df = 5, p-value < 2.2e-16## year Dist_Ecotone Dist_Littoral prop_tmoins1 year2
## 1.091833e+06 1.004869e+00 1.023269e+00 1.018856e+00 1.091827e+062.5.4 Belette - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60716,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -26.32505regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-102
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-103
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dnst_Cultures)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dnst_Cultures, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 11.774340 2.057378 5.723 1.18e-08 ***
## year -0.005760 0.001019 -5.652 1.78e-08 ***
## Dnst_Cultures -0.203133 0.046380 -4.380 1.24e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.04375811)
##
## Null deviance: 105.77 on 2369 degrees of freedom
## Residual deviance: 103.58 on 2367 degrees of freedom
## AIC: -685.15
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-105
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 1.1352, df = 1, p-value = 0.2867##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 51.315, df = 2, p-value = 7.197e-12## year Dnst_Cultures
## 1.000453 1.0004532.6 Fouine
2.6.1 Fouine - Sanglier
bdd_reg <- tab_glm(Donnees,
espece_interet = 60674,
espece_benchmark = 60981)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -286.2294regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-107
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-108
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Ind_Diversite + Dist_Ecotone + Dist_Littoral + Dist_Eau + prop_tmoins1 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Ind_Diversite + Dist_Ecotone +
## Dist_Littoral + Dist_Eau + prop_tmoins1 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.232e+04 3.101e+03 3.973 7.52e-05 ***
## year -1.216e+01 3.073e+00 -3.958 8.01e-05 ***
## Ind_Diversite 4.222e-01 8.005e-02 5.274 1.59e-07 ***
## Dist_Ecotone 5.721e-03 1.075e-03 5.320 1.24e-07 ***
## Dist_Littoral 1.870e-03 7.552e-04 2.476 0.0134 *
## Dist_Eau 1.233e-03 2.568e-04 4.801 1.78e-06 ***
## prop_tmoins1 1.297e-01 2.658e-02 4.879 1.21e-06 ***
## year2 3.002e-03 7.615e-04 3.942 8.55e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.138528)
##
## Null deviance: 218.82 on 1192 degrees of freedom
## Residual deviance: 164.16 on 1185 degrees of freedom
## AIC: 1037.4
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-110
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 9.0779, df = 1, p-value = 0.002587##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 21.486, df = 7, p-value = 0.003113## year Ind_Diversite Dist_Ecotone Dist_Littoral Dist_Eau prop_tmoins1
## 1.150276e+06 1.664282e+00 1.008340e+00 1.624606e+00 1.136317e+00 1.094612e+00
## year2
## 1.150292e+062.6.2 Fouine - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60674,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -503.8788regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-112
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-113
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.388516 0.049502 7.849 6.95e-15 ***
## year2011 -0.046293 0.049280 -0.939 0.347655
## year2012 -0.059439 0.049158 -1.209 0.226758
## year2013 -0.110814 0.049573 -2.235 0.025509 *
## year2014 0.129941 0.048732 2.666 0.007730 **
## year2015 0.238480 0.050326 4.739 2.31e-06 ***
## year2016 0.100554 0.050920 1.975 0.048441 *
## year2017 0.100319 0.051022 1.966 0.049422 *
## year2018 0.069957 0.051747 1.352 0.176564
## year2019 0.124153 0.053136 2.337 0.019567 *
## year2020 0.154598 0.054006 2.863 0.004247 **
## year2021 0.188909 0.052948 3.568 0.000369 ***
## year2022 0.092344 0.053519 1.725 0.084611 .
## year2023 0.010613 0.055811 0.190 0.849206
## year2024 -0.140323 0.055283 -2.538 0.011219 *
## year2025 -0.096578 0.067398 -1.433 0.152034
## X_10km 0.078249 0.009807 7.979 2.52e-15 ***
## prop_tmoins1 0.228099 0.024159 9.442 < 2e-16 ***
## prop_tmoins2 0.177240 0.025804 6.869 8.73e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1332842)
##
## Null deviance: 351.20 on 1932 degrees of freedom
## Residual deviance: 255.11 on 1914 degrees of freedom
## AIC: 1611
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-115
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 3.8186, df = 1, p-value = 0.05069##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 102.5, df = 18, p-value = 7.702e-14## GVIF Df GVIF^(1/(2*Df))
## year -0.2306717 15 NaN
## X_10km 0.1760049 1 0.4195294
## prop_tmoins1 0.1270890 1 0.3564954
## prop_tmoins2 0.1428323 1 0.37793162.6.3 Fouine - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60674,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -244.9157regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-117
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-118
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dist_Ecotone + Dist_Littoral + Dist_Eau + prop_tmoins1 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dist_Ecotone +
## Dist_Littoral + Dist_Eau + prop_tmoins1 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.646e+04 2.535e+03 6.493 1.15e-10 ***
## year -1.628e+01 2.512e+00 -6.480 1.25e-10 ***
## X_10km 6.516e-02 1.366e-02 4.771 2.01e-06 ***
## Dist_Ecotone 4.672e-03 9.145e-04 5.109 3.67e-07 ***
## Dist_Littoral -2.630e-03 6.574e-04 -4.000 6.65e-05 ***
## Dist_Eau 8.523e-04 2.023e-04 4.213 2.68e-05 ***
## prop_tmoins1 1.317e-01 2.596e-02 5.073 4.42e-07 ***
## year2 4.026e-03 6.224e-04 6.467 1.36e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1218901)
##
## Null deviance: 216.52 on 1448 degrees of freedom
## Residual deviance: 175.64 on 1441 degrees of freedom
## AIC: 1072.4
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-120
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 3.154, df = 1, p-value = 0.07574##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 68.39, df = 7, p-value = 3.123e-12## year X_10km Dist_Ecotone Dist_Littoral Dist_Eau prop_tmoins1
## 1.081407e+06 1.720307e+00 1.017864e+00 1.685360e+00 1.038811e+00 1.059038e+00
## year2
## 1.081403e+062.6.4 Fouine - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60674,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -224.37regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-122
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-123
reg <- glm(data = bdd,
formula = proportion_interet ~ X_10km + Ind_Diversite + Dnst_Cultures + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ X_10km + Ind_Diversite + Dnst_Cultures +
## prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.252e+01 1.330e+00 9.415 < 2e-16 ***
## X_10km 6.137e-02 6.786e-03 9.043 < 2e-16 ***
## Ind_Diversite 2.552e-01 5.056e-02 5.047 4.82e-07 ***
## Dnst_Cultures 2.662e-01 9.068e-02 2.935 0.00337 **
## prop_tmoins1 1.146e-01 2.050e-02 5.590 2.53e-08 ***
## prop_tmoins2 7.062e-02 2.121e-02 3.330 0.00088 ***
## year2 -3.153e-06 3.251e-07 -9.699 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.07435608)
##
## Null deviance: 202.30 on 2432 degrees of freedom
## Residual deviance: 180.39 on 2426 degrees of freedom
## AIC: 590.45
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-125
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 2.8128, df = 1, p-value = 0.09352##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 147.28, df = 6, p-value < 2.2e-16## X_10km Ind_Diversite Dnst_Cultures prop_tmoins1 prop_tmoins2 year2
## 1.274696 2.324565 2.275947 1.093374 1.093169 1.0089722.7 Temoins
2.7.1 Sanglier - Renard
bdd_reg <- tab_glm(Donnees,
espece_interet = 60981,
espece_benchmark = 60585)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -1231.308regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-127
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-128
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.135590 0.043534 3.115 0.001869 **
## year2011 0.014981 0.044179 0.339 0.734581
## year2012 -0.008607 0.044159 -0.195 0.845489
## year2013 0.006837 0.044239 0.155 0.877193
## year2014 0.242783 0.043544 5.576 2.81e-08 ***
## year2015 0.261279 0.045780 5.707 1.33e-08 ***
## year2016 0.160117 0.045384 3.528 0.000428 ***
## year2017 0.177714 0.045402 3.914 9.38e-05 ***
## year2018 0.166615 0.045562 3.657 0.000262 ***
## year2019 0.193583 0.046770 4.139 3.64e-05 ***
## year2020 0.295662 0.046869 6.308 3.49e-10 ***
## year2021 0.241756 0.047290 5.112 3.50e-07 ***
## year2022 0.175833 0.046852 3.753 0.000180 ***
## year2023 0.188433 0.047360 3.979 7.18e-05 ***
## year2024 0.159521 0.046429 3.436 0.000603 ***
## year2025 -0.130575 0.059639 -2.189 0.028685 *
## X_10km 0.043284 0.008048 5.378 8.44e-08 ***
## prop_tmoins1 0.387182 0.023994 16.137 < 2e-16 ***
## prop_tmoins2 0.266607 0.025546 10.436 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.09825233)
##
## Null deviance: 378.29 on 1951 degrees of freedom
## Residual deviance: 189.92 on 1933 degrees of freedom
## AIC: 1031.4
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-130
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 12.14, df = 1, p-value = 0.0004935##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 199.33, df = 18, p-value < 2.2e-16## GVIF Df GVIF^(1/(2*Df))
## year 1.7085879 15 1.018016
## X_10km -0.6928078 1 NaN
## prop_tmoins1 -0.7546431 1 NaN
## prop_tmoins2 -0.7989216 1 NaN2.7.2 Sanglier - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60981,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -67.916regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-132
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-133
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + Dist_Littoral + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + Dist_Littoral +
## prop_tmoins1 + prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.2293953 0.1462659 1.568 0.11705
## year2011 0.2559816 0.1626899 1.573 0.11587
## year2012 0.0572882 0.1546126 0.371 0.71105
## year2013 0.0879175 0.1513886 0.581 0.56152
## year2014 0.3246888 0.1415703 2.293 0.02198 *
## year2015 0.2309593 0.1420030 1.626 0.10411
## year2016 0.1465902 0.1418549 1.033 0.30162
## year2017 0.1994991 0.1417965 1.407 0.15969
## year2018 0.2184730 0.1415080 1.544 0.12287
## year2019 0.1958139 0.1419908 1.379 0.16812
## year2020 0.2513929 0.1415922 1.775 0.07606 .
## year2021 0.2095540 0.1414799 1.481 0.13881
## year2022 0.3776144 0.1419111 2.661 0.00789 **
## year2023 0.2343846 0.1421339 1.649 0.09939 .
## year2024 0.2939227 0.1419569 2.071 0.03861 *
## year2025 0.2404299 0.1566266 1.535 0.12502
## X_10km 0.0609917 0.0130918 4.659 3.52e-06 ***
## Dist_Littoral -0.0025615 0.0006317 -4.055 5.32e-05 ***
## prop_tmoins1 0.1370304 0.0309400 4.429 1.03e-05 ***
## prop_tmoins2 0.1025316 0.0328218 3.124 0.00183 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.0954693)
##
## Null deviance: 134.98 on 1282 degrees of freedom
## Residual deviance: 120.58 on 1263 degrees of freedom
## AIC: 649.14
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-135
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 2.3066, df = 1, p-value = 0.1288##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 65.982, df = 19, p-value = 4.225e-07## GVIF Df GVIF^(1/(2*Df))
## year 0.02721716 15 0.8868049
## X_10km 5.64291270 1 2.3754816
## Dist_Littoral 4.81510038 1 2.1943337
## prop_tmoins1 0.39786775 1 0.6307676
## prop_tmoins2 113.91843576 1 10.67325802.7.3 Sanglier - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60981,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -535.5606regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-137
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-138
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.113462 0.034434 3.295 0.000999 ***
## year2011 0.035740 0.038711 0.923 0.355968
## year2012 0.011575 0.037940 0.305 0.760328
## year2013 0.019758 0.037513 0.527 0.598455
## year2014 0.155268 0.035476 4.377 1.26e-05 ***
## year2015 0.081360 0.036047 2.257 0.024095 *
## year2016 0.079757 0.035649 2.237 0.025361 *
## year2017 0.119001 0.035705 3.333 0.000873 ***
## year2018 0.144145 0.035421 4.069 4.87e-05 ***
## year2019 0.055526 0.035164 1.579 0.114460
## year2020 0.137519 0.035421 3.882 0.000106 ***
## year2021 0.119856 0.035385 3.387 0.000718 ***
## year2022 0.092705 0.035098 2.641 0.008314 **
## year2023 0.128777 0.035574 3.620 0.000301 ***
## year2024 0.074951 0.035028 2.140 0.032481 *
## year2025 -0.044103 0.037384 -1.180 0.238226
## X_10km 0.038429 0.005203 7.386 2.09e-13 ***
## prop_tmoins1 0.259656 0.021406 12.130 < 2e-16 ***
## prop_tmoins2 0.188531 0.022884 8.239 2.87e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.04938447)
##
## Null deviance: 151.10 on 2344 degrees of freedom
## Residual deviance: 114.87 on 2326 degrees of freedom
## AIC: -378.3
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-140
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 6.2571, df = 1, p-value = 0.01237##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 115.65, df = 18, p-value = 2.768e-16## GVIF Df GVIF^(1/(2*Df))
## year -0.367008173 15 NaN
## X_10km -0.006618221 1 NaN
## prop_tmoins1 -0.017044621 1 NaN
## prop_tmoins2 -0.010353522 1 NaN2.7.4 Renard - Chevreuil
bdd_reg <- tab_glm(Donnees,
espece_interet = 60585,
espece_benchmark = 61057)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -1147.459regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-142
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-143
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dist_Ecotone + Dist_Littoral +
## prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.627e+04 1.838e+03 19.736 < 2e-16 ***
## year -3.592e+01 1.822e+00 -19.715 < 2e-16 ***
## Dist_Ecotone 2.743e-03 5.280e-04 5.195 2.24e-07 ***
## Dist_Littoral -2.350e-03 4.323e-04 -5.436 6.06e-08 ***
## prop_tmoins1 3.158e-01 1.916e-02 16.480 < 2e-16 ***
## prop_tmoins2 1.014e-01 1.950e-02 5.200 2.17e-07 ***
## year2 8.893e-03 4.515e-04 19.695 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1205656)
##
## Null deviance: 455.42 on 2188 degrees of freedom
## Residual deviance: 263.07 on 2182 degrees of freedom
## AIC: 1590.1
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-145
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 19.932, df = 1, p-value = 8.026e-06##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 96.343, df = 6, p-value < 2.2e-16## year Dist_Ecotone Dist_Littoral prop_tmoins1 prop_tmoins2 year2
## 1.097773e+06 1.021077e+00 1.056587e+00 1.415572e+00 1.430218e+00 1.097745e+062.7.5 Renard - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 60585,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "quantitative"
##
## $min_bic
## [1] -624.2308regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-147
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-148
reg <- glm(data = bdd,
formula = proportion_interet ~ year + Dnst_Cultures + Dist_Eau + prop_tmoins1 + prop_tmoins2 + year2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + Dnst_Cultures + Dist_Eau +
## prop_tmoins1 + prop_tmoins2 + year2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.710e+04 1.580e+03 10.826 < 2e-16 ***
## year -1.692e+01 1.566e+00 -10.805 < 2e-16 ***
## Dnst_Cultures -3.996e-01 7.271e-02 -5.496 4.22e-08 ***
## Dist_Eau 4.381e-04 1.305e-04 3.355 0.000803 ***
## prop_tmoins1 1.057e-01 1.769e-02 5.977 2.56e-09 ***
## prop_tmoins2 6.444e-02 1.807e-02 3.567 0.000367 ***
## year2 4.185e-03 3.881e-04 10.783 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.1266005)
##
## Null deviance: 458.87 on 2865 degrees of freedom
## Residual deviance: 361.95 on 2859 degrees of freedom
## AIC: 2219.1
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-150
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 10.734, df = 1, p-value = 0.001052##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 74.369, df = 6, p-value = 5.176e-14## year Dnst_Cultures Dist_Eau prop_tmoins1 prop_tmoins2 year2
## 1.078080e+06 1.032443e+00 1.032468e+00 1.104598e+00 1.111186e+00 1.078105e+062.7.6 Chevreuil - Blaireau
bdd_reg <- tab_glm(Donnees,
espece_interet = 61057,
espece_benchmark = 60636)
liste <- choisi_forme_year(bdd = bdd_reg)## $choice
## [1] "qualitative"
##
## $min_bic
## [1] -623.9254regfit <- liste[1]$regfit
bdd <- liste[2]$bddreg.summary <- summary(regfit); plot_regsubsets(reg.summary)
plot of chunk unnamed-chunk-152
par(mfrow=c(1 ,1)) ; plot(regfit, scale ="bic")
plot of chunk unnamed-chunk-153
reg <- glm(data = bdd,
formula = proportion_interet ~ year + X_10km + prop_tmoins1 + prop_tmoins2)summary(reg); autoplot(reg)##
## Call:
## glm(formula = proportion_interet ~ year + X_10km + prop_tmoins1 +
## prop_tmoins2, data = bdd)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.201699 0.041696 4.837 1.4e-06 ***
## year2011 0.002699 0.047090 0.057 0.954292
## year2012 0.026320 0.045647 0.577 0.564267
## year2013 0.037639 0.045199 0.833 0.405070
## year2014 0.149116 0.043029 3.465 0.000539 ***
## year2015 0.121364 0.043695 2.778 0.005520 **
## year2016 0.124997 0.043515 2.872 0.004108 **
## year2017 0.107340 0.043632 2.460 0.013959 *
## year2018 0.134913 0.043426 3.107 0.001914 **
## year2019 0.040537 0.042989 0.943 0.345789
## year2020 0.101893 0.043336 2.351 0.018792 *
## year2021 0.141254 0.043031 3.283 0.001043 **
## year2022 -0.017666 0.042892 -0.412 0.680469
## year2023 0.128744 0.043423 2.965 0.003058 **
## year2024 0.007942 0.042680 0.186 0.852403
## year2025 -0.131530 0.045667 -2.880 0.004009 **
## X_10km 0.041494 0.006286 6.601 5.0e-11 ***
## prop_tmoins1 0.284694 0.020749 13.721 < 2e-16 ***
## prop_tmoins2 0.184130 0.021765 8.460 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 0.07749643)
##
## Null deviance: 252.40 on 2433 degrees of freedom
## Residual deviance: 187.15 on 2415 degrees of freedom
## AIC: 703.31
##
## Number of Fisher Scoring iterations: 2
plot of chunk unnamed-chunk-155
bgtest(reg); bptest(reg); car::vif(reg)##
## Breusch-Godfrey test for serial correlation of order up to 1
##
## data: reg
## LM test = 5.8301, df = 1, p-value = 0.01575##
## studentized Breusch-Pagan test
##
## data: reg
## BP = 59.363, df = 18, p-value = 2.592e-06## GVIF Df GVIF^(1/(2*Df))
## year 0.7688369 15 0.9912757
## X_10km -0.1951470 1 NaN
## prop_tmoins1 -0.2983212 1 NaN
## prop_tmoins2 -0.3295191 1 NaN3 Mortalite routieres - DIRO uniquement
Il me manque la colonne bdd_originale… Ou est elle passe?