ID
int64 1
1.07k
| Comment
stringlengths 8
1.13k
| Code
stringlengths 10
4.28k
| Label
stringclasses 4
values | Source
stringlengths 21
21
| File
stringlengths 4
82
|
|---|---|---|---|---|---|
101 | Network analysis using EBICglasso with the Germanspeaking sample including the openmindedness scale | gr2<- list('Deliberate MW'=c(1:4), 'Spontaneous MW'=c(5:8), 'Boredom'=c(9:16), 'Open-Mindedness'=c(17:22)) names2<-c("I allow my thoughts to wander on purpose", "I enjoy mind-wandering", "I find mind-wandering is a good way to cope with boredom", "I allow myself to get absorbed in pleasant fantasy", "I find my thoughts wandering spontaneously", "When I mind-wander my thoughts tend to be pulled from topic to topic", "It feels like I donâÂÂt have control over when my mind wanders", "I mind wander even when IâÂÂm supposed to be doing something else", "I often find myself at âÂÂloose ends,â not knowing what to do", "I find it hard to entertain myself", "Many things I have to do are repetitive and monotonous", "It takes more stimulation to get me going than most people", "I donâÂÂt feel motivated by most things that I do", "In most situations, it is hard for me to find something to do or see to keep me interested", "Much of the time, I just sit around doing nothing", "Unless I am doing something exciting, even dangerous, I feel half-dead and dull", "Has few artistic interests", "Is complex, a deep thinker", "Is original, comes up with new ideas", "Is fascinated by art, music, or literature", "Has little interest in abstract ideas", "Has little creativity") n2<-estimateNetwork(DataGerman, default= "EBICglasso") plot(n2, groups = gr2, nodeNames = names2, legend.cex=.35) centrality_auto(n2, weighted = TRUE, signed = TRUE) centralityPlot(n2, include =c("Betweenness","Closeness", "Strength")) print(n2) | Statistical Modeling | https://osf.io/tg3fq/ | syntax_SDMWS&SBPS.R |
102 | Extract the vectors of analysed characteristics of Partner and Nonbiological father from the dataset. | text<-paste("trait_p<-data$Partner_",char,sep="") eval(parse(text=text)) text<-paste("trait_f<-data$Nonbiol_",char,sep="") eval(parse(text=text)) text<-paste("trait_pb<-dbiol$Partner_",char,sep="") eval(parse(text=text)) text<-paste("trait_b<-dbiol$Biol_",char,sep="") eval(parse(text=text)) | Data Variable | https://osf.io/greqt/ | functions2.R |
103 | Cohen's d First create the equivalent arrangement for the stimation of differences if the father is biologicla and present | muB<-rep(mean(diffb,na.rm=T),15) sdB<-rep(sd(diffb,na.rm=T),15) nB<-rep(length(diffb[!is.na(diffb)]),15) | Statistical Test | https://osf.io/greqt/ | functions2.R |
104 | plots also when there is no interaction and also squared terms (potentially involved in interactions) create data frame storing the observed mean values per combination of the values of the grided covariate and factor levels in the data create vectors denoting the grid cell center values for the two covariate: | if(!is.na(grid.resol)){ xvar=seq(min(plot.data[,covariate]), max(plot.data[,covariate]), length.out=grid.resol) bin.x=cut(x=plot.data[,covariate], breaks=xvar, include.lowest=T, labels=F) bin.x=min(xvar)+diff(xvar[1:2])/2+(bin.x-1)*diff(xvar[1:2]) }else{ bin.x=plot.data[, covariate] } | Visualization | https://osf.io/vjeb3/ | draw.2.w.int.bw.1.cov.and.1.fac.r |
105 | remove rows where wkl_uuid is NA: | VS <- VS[!is.na(VS$wkl_uuid), ] | Data Variable | https://osf.io/w7pjy/ | format_captWKLquality.R |
106 | Generate data assuming an underlying VAR(1) process Function to generate data assuming an underlying VAR(1) process the input is N the number of samples, T.days and T.beeps are the number of days and the number of beeps ESM is conducted (Total number of assessments is T.days x T.beeps) Psi is a matrix with the fixed regression weights, mu is a vector with the fixed intercepts, var.Psi is variance of the random regression weights (i.e., assumed to be equal), For each participant, it is checked whether the matrix Psi conforms the assumption of stationary time series with the absolute value of the maximum eigenvalue smaller than 1 The distribution of the random effects resembles a truncated multivariate normal distribution. This function simulates data from a VAR(1) model | Data.VAR.Fixed = function(N,T,Psi,cor.Sigma){ p = ncol(Psi) | Statistical Modeling | https://osf.io/rs6un/ | Data.VAR.Fixed.R |
107 | remove layers with gtype_class "MF" (only three distinct MF layers in data set!) | toDiscard <- which(CM$gtype_class == "MF") if (length(toDiscard) > 0) CM <- CM[-toDiscard, ] | Data Variable | https://osf.io/w7pjy/ | format_confusionMatrix.R |
108 | Testing participant heterogeneity chisquare test: | testHetChi(freq = "data/data_retrieval.csv", tree = c("E","E","E", "U","U","U", "N","N","N") ) testHetChi(freq = "data/data_encoding.csv", tree = c("E","E","E", "U","U","U", "N","N","N") ) | Statistical Test | https://osf.io/s82bw/ | 01_TreeBUGS_with_csv_files.R |
109 | Fitting a betaMPT model | m.retrieval.beta <- betaMPT(eqnfile="model/2htsm.eqn", data = "data/data_retrieval.csv", restrictions = "model/restrictions.txt", modelfilename = "results/2htsm_betaMPT.jags", transformedParameters = list("deltaDd=D1-d1"), parEstFile = "results/results_retrieval_betaMPT.txt", n.chain = 4, n.iter = 50000, n.adapt = 10000, n.burnin = 10000, n.thin = 10, ppp = 5000, dic = TRUE ) summary(m.retrieval.beta) | Statistical Modeling | https://osf.io/s82bw/ | 01_TreeBUGS_with_csv_files.R |
110 | Withinsubject tests Example: twohigh threshold model (2HTM, included in TreeBUGS) | htm <- system.file("MPTmodels/2htm.eqn", package="TreeBUGS") | Statistical Modeling | https://osf.io/s82bw/ | 01_TreeBUGS_with_csv_files.R |
111 | Plot ranked correlations | cor_dat <- corr_cross(ogdatasub, max_pvalue = 0.05, # display only significant correlations (at 5% level) pvalue =T, plot = F, top = 10 # display top 10 correlations (by correlation coefficient) ) |> arrange(abs(corr)) |> mutate(lab = paste(key, mix, sep = " + "), lab = factor(lab)) cor_dat |> ggplot(aes(y = lab, x = abs(corr), fill = as.factor(sign(corr)))) + geom_col() + scale_fill_manual(values = c("#E64B3D", "#0092B2")) + labs(title="", subtitle = "", caption="Note: Red bars indicate a negative relationship") + geom_text(aes(label = signif(.data$corr, 2)), colour = "#FFFFFF", size = 3.5, hjust = 1.1) + scale_x_continuous(limits=c(0, .4), position="top") + scale_y_discrete(labels=c("fst + cst" = "Difficult Financial Situation & Cost of Food", "rsp + spp" = "Loved Ones' Needs & Lack of Support", "pdd + ppd" = "Positive Experiences with Veg*nism & Positive Elements Post-Diet", "eft + prc" = "Meal Complexity & Need for Specialized Equipment or Planning", "fds + oth" = "Discontent and Cravings for Animal Products & Other Obstacles", "rsp + eft" = "Loved Ones' Needs & Meal Complexity", "fph + cst" = "Challenging Purchase Requirements & Cost of Food", "avl + fph" = "Limited Access to Veg*n Options & Challenging Purchase Requirements", "med + con" = "Professional Medical Advice & Difficulty Managing Health", "con + com" = "Difficulty Managing Health & Commitment Difficulties", "fcn + prc" = "Lack of Veg*n Knowledge & Need for Specialized Equipment or Planning")) + theme_minimal() + theme(axis.title = element_blank(), legend.position = "none") cor_dat ggsave("Correlations.png", width=8, height=6) | Visualization | https://osf.io/q2zrp/ | Graphs.R |
112 | Predictive performance plot ( model size selection plot) | stat_pretty <- setNames(nm = proj_evalstats) stat_pretty <- toupper(stat_pretty) stopifnot(identical(proj_evalstats, "mlpd")) ggeval <- plot(C_cvvs, stats = proj_evalstats, deltas = TRUE, ranking_nterms_max = NA) ggeval <- ggeval + facet_null() ggeval <- ggeval + scale_y_continuous( sec.axis = sec_axis(~ exp(.), name = bquote(Lambda*" "*"GMPD")) ) ggeval <- ggeval + labs(y = bquote(Delta*" "*.(stat_pretty))) print(ggeval) ggsave(file.path("output", out_folder, paste0(plot_prefix, "projpred_search_deltas.jpeg")), width = 7, height = 7 * 0.618) saveRDS(last_plot(), file = file.path("output", out_folder, paste0(plot_prefix, "projpred_search_deltas.rds"))) | Visualization | https://osf.io/emwgp/ | projpred.R |
113 | save attribute level for option i in trial row.id | wide.data[[paste0("opt",i,att.names[k])]][row.id]=feat.ij | Data Variable | https://osf.io/tbczv/ | 01-readData.r |
114 | save position of attribute j | wide.data[[paste0("pos",att.names[k])]][row.id]=j | Data Variable | https://osf.io/tbczv/ | 01-readData.r |
115 | save attribute of position k | wide.data[[paste0("pos",j)]][row.id]=att.names[k] } } } if(y$selected[i]==1) wide.data$response[row.id]=i } | Data Variable | https://osf.io/tbczv/ | 01-readData.r |
116 | 1. recode RT for trials with slow (missed) response as NA | data$rt[data$trialError == " Slow"] <- NA | Data Variable | https://osf.io/tbczv/ | 01-readData.r |
117 | ttests to compare composite 'animal protection behaviors' in people who have vs. have not experienced each (16 types total) | t.test(beh_animalprotect_comp ~ graphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp~ nongraphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ person_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ leaflet_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ news_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ social_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ humaneed_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ documentary_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ book_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ celebrity_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ ad_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ challenge_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ labels_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp~ labeled_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ ndisprotest_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ disprotest_exp_buc, data=data, var.equal = TRUE) | Statistical Test | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
118 | FDR correction for ttests that compared composite 'animal protection behaviors' between people who have and have not experienced each advocacy type | adjusted_pvalues_protectionbehaviors <- data %>% summarise(graphic_exp_buc = t.test(beh_animalprotect_comp ~ graphic_exp_buc, var.equal = TRUE)$p.value, nongraphic_exp_buc = t.test(beh_animalprotect_comp ~ nongraphic_exp_buc, var.equal = TRUE)$p.value, person_exp_buc = t.test(beh_animalprotect_comp ~ person_exp_buc, var.equal = TRUE)$p.value, leaflet_exp_buc = t.test(beh_animalprotect_comp ~ leaflet_exp_buc, var.equal = TRUE)$p.value, news_exp_buc = t.test(beh_animalprotect_comp ~ news_exp_buc, var.equal = TRUE)$p.value, social_exp_buc = t.test(beh_animalprotect_comp ~ social_exp_buc, var.equal = TRUE)$p.value, humaneed_exp_buc = t.test(beh_animalprotect_comp ~ humaneed_exp_buc, var.equal = TRUE)$p.value, documentary_exp_buc = t.test(beh_animalprotect_comp ~ documentary_exp_buc, var.equal = TRUE)$p.value, book_exp_buc = t.test(beh_animalprotect_comp ~ book_exp_buc, var.equal = TRUE)$p.value, celebrity_exp_buc = t.test(beh_animalprotect_comp ~ celebrity_exp_buc, var.equal = TRUE)$p.value, ad_exp_buc = t.test(beh_animalprotect_comp ~ ad_exp_buc, var.equal = TRUE)$p.value, challenge_exp_buc = t.test(beh_animalprotect_comp ~ challenge_exp_buc, var.equal = TRUE)$p.value, labels_exp_buc = t.test(beh_animalprotect_comp ~ labels_exp_buc, var.equal = TRUE)$p.value, labeled_exp_buc = t.test(beh_animalprotect_comp ~ labeled_exp_buc, var.equal = TRUE)$p.value, ndisprotest_exp_buc = t.test(beh_animalprotect_comp ~ ndisprotest_exp_buc, var.equal = TRUE)$p.value, disprotest_exp_buc = t.test(beh_animalprotect_comp ~ disprotest_exp_buc, var.equal = TRUE)$p.value) %>% gather("Advocacy","p_value") %>% mutate(p_fdr = p.adjust(p_value, method = "fdr", n = length(p_value))) %>% print() | Statistical Test | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
119 | ttests to compare composite 'animal consumer behaviors' in people who have vs. have not experienced each (16 types total) | t.test(beh_consumer_comp ~ graphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp~ nongraphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ person_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ leaflet_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ news_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ social_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ humaneed_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ documentary_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ book_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ celebrity_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ ad_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ challenge_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ labels_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ labeled_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ ndisprotest_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ disprotest_exp_buc, data=data, var.equal = TRUE) | Statistical Test | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
120 | prepare animal protection behaviors for stacked graph | mean_CI_protect <- data %>% group_by(advocacytype, experienced) %>% summarize(n = n(), protect_mean = mean(beh_animalprotect_comp), protect_lci = t.test(beh_animalprotect_comp, conf.level = 0.95)$conf.int[1], protect_uci = t.test(beh_animalprotect_comp, conf.level = 0.95)$conf.int[2]) | Visualization | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
121 | add asterisks to advocacy types that had sig different scores between experienced vs. not experienced from ttests | mean_CI_protect <- mean_CI_protect %>% mutate(advocacytype = case_when(advocacytype == "ad" ~ "Anuncio o Valla Publicitaria*", advocacytype == "book" ~ "Libro*", advocacytype == "celebrity" ~ "Celebridad*", advocacytype == "challenge" ~ "Reto de Evitar la Carne*", advocacytype == "disprotest" ~ "Protesta Disruptiva*", advocacytype == "graphic" ~ "Video Gráfico*", advocacytype == "humaneed" ~ "Educación en el Aula de Clases*", advocacytype == "labeled" ~ "Información Educativa Sobre las Etiquetas de Bienestar*", advocacytype == "labels" ~ "Etiquetas Vegano/De Base Vegetal*", advocacytype == "ndisprotest" ~ "Protesta No Disruptiva*", advocacytype == "news" ~ "ArtÃculo de Noticias*", advocacytype == "nongraphic" ~ "Video No Gráfico*", advocacytype == "person" ~ "Divulgación Boca a Boca *", advocacytype == "social" ~"Publicación en Redes Sociales o en un Blog*", advocacytype == "documentary" ~ "Documental*", advocacytype == "leaflet" ~ "Folleto o Volante*")) mean_CI_consume <- mean_CI_consume %>% mutate(advocacytype = case_when(advocacytype == "ad" ~ "Anuncio o Valla Publicitaria*", advocacytype == "book" ~ "Libro*", advocacytype == "celebrity" ~ "Celebridad*", advocacytype == "challenge" ~ "Reto de Evitar la Carne*", advocacytype == "disprotest" ~ "Protesta Disruptiva*", advocacytype == "graphic" ~ "Video Gráfico*", advocacytype == "humaneed" ~ "Educación en el Aula de Clases*", advocacytype == "labeled" ~ "Información Educativa Sobre las Etiquetas de Bienestar*", advocacytype == "labels" ~ "Etiquetas Vegano/De Base Vegetal*", advocacytype == "ndisprotest" ~ "Protesta No Disruptiva*", advocacytype == "news" ~ "ArtÃculo de Noticias*", advocacytype == "nongraphic" ~ "Video No Gráfico*", advocacytype == "person" ~ "Divulgación Boca a Boca *", advocacytype == "social" ~"Publicación en Redes Sociales o en un Blog*", advocacytype == "documentary" ~ "Documental*", advocacytype == "leaflet" ~ "Folleto o Volante*")) | Statistical Test | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
122 | factorize advocacy type for graph | mean_CI_protect <- mean_CI_protect %>% mutate(advocacytype = factor(advocacytype, bar_order)) mean_CI_consume <- mean_CI_consume %>% mutate(advocacytype = factor(advocacytype, bar_order)) | Data Variable | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
123 | grouped bar graph for animal protection behaviors | ggplot(mean_CI_protect, aes(x= advocacytype, fill=experienced, y=protect_mean)) + geom_col(width = 0.8, position="dodge") + coord_flip(ylim=c(1,5)) + geom_errorbar(aes(x= advocacytype, ymin = protect_lci, ymax = protect_uci), width=0.4, colour = "black", position=position_dodge(.8)) + geom_text(aes(label = format(round(protect_mean, 1)), y = protect_uci), hjust = -0.2,# nudge_y = 2, size = 3, position = position_dodge(width = 1)) + labs(y= "Puntuación Media del Comportamiento de Protección de los Animales", x = "Tipo de Defensa", caption = "Un asterisco (*) indica que hubo una diferencia estadÃsticamente significativa entre los grupos (todos los ps < 0,05) después de haber corregido mediante FDR. Para más detalles sobre cómo se llevaron a cabo estos análisis, véase el Material Complementario.") + theme(legend.title = element_blank(), legend.position = "bottom", panel.background = element_rect("white"), panel.border = element_rect(fill = NA), panel.grid.major.x = element_line("grey")) + scale_fill_manual(values = c("Experimentado" = "#c47020", "No Experimentado" = "#F68D29")) | Visualization | https://osf.io/3aryn/ | 7AnimalProtectionandConsumerBehaviors_Spanish.R |
124 | We create a list of parameters to visualize and fill the labels etc to the .txt file in any table editor (Excel) | write.table(data.frame(nam=rep.par2),"params_empty.txt",row.names=F,sep="\t") | Visualization | https://osf.io/fr5ed/ | 03_posterior_visualization_country_contrast.R |
125 | plot contrasts estimates | plot(NULL,ylim=c(max(dec$y)+const1+const2,min(dec$y)-const1),type="n",xaxs="i",yaxs="i",xaxt="n",yaxt="n",xlim=xlims[[block]],bty="n") abline(h=dec$y,col=col.grid,lty=1,lwd=lwd.grid) segments(axes[[block]],max(dec$y)+const1,axes[[block]],max(dec$y)+tic+const1,lwd=lwd.ax,col=col.ax) text(axes[[block]],max(dec$y)+tic+const1+ofs,labels=axes[[block]],col=col.ax,cex=0.9,font=2) lines(range(axes[[block]]),rep(max(dec$y)+const1,2),lwd=lwd.ax,col=col.ax) segments(axes[[block]],min(dec$y)-const1,axes[[block]],max(dec$y)+const1,lwd=lwd.v,col=col.ax,lty=3) segments(0,min(dec$y)-const1,0,max(dec$y)+const1,lwd=lwd.v,col=col.ax,lty=1) | Visualization | https://osf.io/fr5ed/ | 03_posterior_visualization_country_contrast.R |
126 | Draw density polygons density areas are scaled within each block. | area<-0.25*diff(xlims[[block]]) for(i in 1:nrow(dec)){ thispost<-postsc[,dec$n[i]] dens<-density(thispost) polX<-c(dens$x,rev(dens$x)) polY<-c(dens$y,rev(-dens$y)) ar1<-abs(polyarea(polX,polY)) perc<-area/ar1 polygon(polX,polY*perc+dei$y[i],col=dec$hex[i],border=col.pol,lwd=lwd.pol) | Visualization | https://osf.io/fr5ed/ | 03_posterior_visualization_country_contrast.R |
127 | correlations between csd_mean and averaged csds for Big Five traits (mentioned in the text in Appendix A): | psych::corr.test( df2$n_csd,df2$csd_mean_n ) psych::corr.test( df2$e_csd,df2$csd_mean_e ) psych::corr.test( df2$o_csd,df2$csd_mean_o ) psych::corr.test( df2$a_csd,df2$csd_mean_a ) psych::corr.test( df2$c_csd,df2$csd_mean_c ) | Data Variable | https://osf.io/tajd9/ | Flip_MainDataAnalyses.R |
128 | correlations between variability and wellbeing measures: | names_var <- c("csd_mean_n","csd_mean_e","csd_mean_o","csd_mean_a","csd_mean_c", "sccs","csd_nob","simpson_csd","rses","swls","pa","na") psych::corr.test( df2[,names_var] ) | Statistical Modeling | https://osf.io/tajd9/ | Flip_MainDataAnalyses.R |
129 | catch situation where Sx is missing so we cannot assess response to treatment | if ( nrow( thisSx ) < 1 ) { thisRow$status.sx <- NA noFurtherFlag <- 1 ## tell rest of loop to not bother thisRow$missing.Sx <- 1 } | Data Variable | https://osf.io/5y27d/ | tabulate_cases.R |
130 | Descriptive analysis: counting occurrence of bodybased units and coded themes. Step 1. Analyse frequencies of bodybased units of measure Ensure data is in character format | df$`Body dimension` <- as.character(df$`Body dimension`) | Data Variable | https://osf.io/fegvr/ | Analysis.R |
131 | Repeating confirmatory analyses with extra exclusions: in addition to excluding participants who did not complete the study, and excluding timeuntilguess and arithmeticsolvingtimes more than 5 standard deviations away from the mean, this analysis also excludes participants who removed 0 tiles when guessing. Exclude participants who did not complete study | d.conf_1_2.complete <- d.conf_1_2.b[complete.cases(d.conf_1_2.b),] d.conf_3_math.complete <- d.conf_3_math.b[complete.cases(d.conf_3_math.b),] | Data Variable | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
132 | remove rows with timeuntilguess values that are more than 5 standard deviations away from the mean | agg <- aggregate(ElapsedTime_Guess ~ Guess_Number + ID_Player + Effort + Competition, data=d.conf_1_2.complete, FUN = mean) sd5.times <- mean(agg$ElapsedTime_Guess) + (5 * sd(agg$ElapsedTime_Guess)) nrow(agg[agg$ElapsedTime_Guess > sd5.times,]) | Statistical Test | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
133 | remove rows with arithmeticsolvingtimes that are more than 5 standard deviations away from the mean | sd5.times <- mean(d.conf_3_math.complete$ElapsedTime_Math) + (5 * sd(d.conf_3_math.complete$ElapsedTime_Math)) nrow(d.conf_3_math.complete[d.conf_3_math.complete$ElapsedTime_Math > sd5.times,]) | Statistical Test | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
134 | Exclude participants who revealed 0 tiles | d.conf_1_2.complete_no0 <- d.conf_1_2.complete[d.conf_1_2.complete$TilesRevealed > 0,] d.conf_3_math.complete_no0 <- d.conf_3_math.complete[d.conf_3_math.complete$TilesRevealed > 0,] | Data Variable | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
135 | compare simpler models with sex and sex*effort | df <- compare(m_accuracy_guess_nointer, m_accuracy_componly, m_accuracy_sex, m_accuracy_sex_effort_inter, m_accuracy_sex_effort_inter_CEinter) df <- round(df@output, 2) df %>% kable(caption = "ACCURACY") %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive"), full_width = TRUE) %>% column_spec(1:5, width = "2cm") | Statistical Modeling | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
136 | aggregate data for time elapsed per guess, by first removing duplicate times, then summing up total times per player | d.reward.agg <- aggregate(ElapsedTime_Guess ~ Effort + Competition + Sex + Guess_Number + ID_Player, data = d.reward, FUN = mean) d.reward.agg <- aggregate(ElapsedTime_Guess ~ Effort + Competition + Sex + ID_Player, data = d.reward.agg, FUN = sum) | Data Variable | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
137 | Bayes Factors load additional libraries subset data into separate objects for frequentist analyses | d.math.agg.f <- d.math.agg d.math.agg.f$Competition <- as.factor(d.math.agg.f$Competition) d.math.agg.f$ID_Player <- as.factor(d.math.agg.f$ID_Player) d.math.agg.f$Sex <- as.factor(d.math.agg.f$Sex) d.conf.agg.f <- d.conf.agg d.conf.agg.f$Competition <- as.factor(d.conf.agg.f$Competition) d.conf.agg.f$Effort <- as.factor(d.conf.agg.f$Effort) d.conf.agg.f$ID_Player <- as.factor(d.conf.agg.f$ID_Player) d.conf.agg.f$Sex <- as.factor(d.conf.agg.f$Sex) | Data Variable | https://osf.io/7vbj9/ | Analyses_Exploratory.R |
138 | Plot AUC | ggplot(plot.data, aes(x = Target, y = AUC, color = Algorithm, fill = Algorithm, shape = Algorithm)) + geom_boxplot(width = 0.3,lwd = 1, aes(color = Algorithm, fill = Algorithm), alpha = 0.3, outlier.shape=NA, position=position_dodge(0.5)) + geom_point(position=position_jitterdodge(jitter.width = 0.1, dodge.width = 0.5), size = 0.5) + coord_flip() + geom_hline(yintercept = 0.5, color="black", linetype = "dashed", size = 1) + theme_classic() + theme(text = element_text(size = 18), axis.title.y = element_blank(), legend.position = c(0.85, 0.5)) + scale_color_manual(values = c("#440154", "#5ec962")) + scale_fill_manual(values = c("#440154", "#5ec962")) + ylim(0.25, 1) + scale_x_discrete(labels=rev(c("Sociality" = expression(paste("* ", bold("S"), "ociality")), "Deception" = expression(paste(bold("D"), "eception")), "Negativity" = expression(paste(bold("N"), "egativity")), "Positivity" = expression(paste("* p", bold("O"), "sitivity")), "Mating" = expression(paste("* ", bold("M"), "ating")), "Adversity" = expression(paste(bold("A"), "dversity")), "Intellectuality" = expression(paste("* ", bold("I"), "ntellect")), "Duty" = expression(paste("* ", bold("D"), "uty"))))) | Visualization | https://osf.io/b7krz/ | 01_ML_bmr_resultsSummary.R |
139 | Step 4*: Visualization of the ROC curves for "core benchmark" for the Online Supplemental Material Duty | p_Duty = autoplot(bmr$clone(deep = TRUE)$filter(task_ids = "Duty", learner_ids = c("lasso_Duty", "rf_Duty")), type = "roc") + theme_classic() + theme(title = element_blank(), legend.position = "none", axis.title = element_text(size = 18), axis.text = element_text(size = 16)) + labs(y = "Sensitivity", x = "1 - Specifity") + geom_abline(intercept = 0, slope = 1, color = "black", linetype = "dotted", size = 0.8) + annotate(geom = "text", x = 0, y = 1, label = "Duty", hjust = 0, vjust = 1, size = 10) png("Figures_Tables/Figures/ROC_Curves/Duty.png", width = 450, height = 400) p_Duty dev.off() | Visualization | https://osf.io/b7krz/ | 01_ML_bmr_resultsSummary.R |
140 | Plot ROC curves and caculate AUC (if desired) | plotIt <- FALSE currAUC <- plotAUC(allPs,allBFs,dt,numStudies,plotIt) #currently does not save the AUCs ai <- ai+1 aucs[ai,1] <- sampleSizes[1] aucs[ai,2] <- effectSizes[2] aucs[ai,3] <- i #runNumber aucs[ai,4] <- currAUC[1] aucs[ai,5] <- currAUC[2] plotIt <- FALSE currAUC <- plotAUC(allPs2,allBFs2,dt,numStudies,plotIt) #currently does not save the AUCs aucs$aucPhacked[ai] <- currAUC[1] aucs$aucBFhacked[ai] <- currAUC[2] if (currAUC[1] != currAUC[2]) { print(paste(effectSizes,i)) print(currAUC) } } #end for i runTimes } #end for ssa allDt <- allDt[-1,] #get rid of initial dummy row that was just used to init allDt | Visualization | https://osf.io/hzncs/ | Witt_SDT_Simulations_OptionalStopping.R |
141 | Stacked plots by SDT outcome Plot by whether hacked or not plot hit, FA, miss, corrRejection for each effect size/sample size combo | ap <- aggregate(hits ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) ap2 <- aggregate(fa ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(ap,ap2,by=c("criterion","sampleSize","effectSizes","withHack")) ap2 <- aggregate(miss ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(hFA,ap2,by=c("criterion","sampleSize","effectSizes","withHack")) ap2 <- aggregate(corrRej ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(hFA,ap2,by=c("criterion","sampleSize","effectSizes","withHack")) hFA$hits <- hFA$hits / numStudies hFA$fa <- hFA$fa / numStudies hFA$corrRej <- hFA$corrRej / numStudies hFA$miss <- hFA$miss / numStudies | Visualization | https://osf.io/hzncs/ | Witt_SDT_Simulations_OptionalStopping.R |
142 | plot BF vs pvalues | plot(log(allPs),log(allBFs),bty="l") | Visualization | https://osf.io/hzncs/ | Witt_SDT_Simulations_OptionalStopping.R |
143 | Create matrix of input variables with populationlevel effects: | C_datMM <- model.matrix( as.formula(paste("~", paste(vpreds_noInt, collapse = " + "))), data = C_dat ) | Data Variable | https://osf.io/emwgp/ | ppfs.R |
144 | Set the latent regression coefficients for booklet 1, 5, 7 and 10 for reading and booklet 4, 6, 9 and 11 for science to 0 | betas <- data.frame(var = c(which(names(con_dat) %in% paste0("bookid.", c(1, 5, 7, 10))), which(names(con_dat) %in% paste0("bookid.", c(4, 6, 9, 11)))), dim = rep(2:3, each = 4), value = 0) betas <- data.frame(var = c(which(names(con_dat) %in% paste0("bookid.", c(1, 5, 7, 10))), which(names(con_dat) %in% paste0("bookid.", c(4, 6, 9, 11)))), dim = rep(2:3, each = 4), value = 0) | Data Variable | https://osf.io/8fzns/ | 4H_PV_helper.R |
145 | Compute latent regression of the latent ability on the conditioning variables (excl. the first column which is the student ID) | latreg <- tam.latreg(likeli, Y = con_dat[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = "Ramsay"), beta.fixed = as.matrix(betas)) } else { latreg <- tam.latreg(likeli, Y = con_dat[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = "Ramsay")) } latreg <- tam.latreg(likeli, Y = con_dat1[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = "Ramsay"), beta.fixed = as.matrix(betas)) } else { latreg <- tam.latreg(likeli, Y = con_dat1[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = "Ramsay")) } | Statistical Modeling | https://osf.io/8fzns/ | 4H_PV_helper.R |
146 | Draw 5 plausible values for each student out of the resulting distribution which is assumed to be normal distributed | pvs <- tam.pv(latreg, nplausible = 5, normal.approx = T, samp.regr = samp.regr.opt) } else { pvs <- tam.pv(latreg_md, nplausible = 5, normal.approx = T, samp.regr = samp.regr.opt) } else { | Statistical Modeling | https://osf.io/8fzns/ | 4H_PV_helper.R |
147 | Extract the regression coefficients of the conditioning variables, because they are fixed for the core domains in the next step at that value | reg.coefs <- cbind(rep(1:dim(latreg$beta)[1], 3), rep(1:3, each = dim(latreg$beta)[1]), c(latreg$beta[, 1], latreg$beta[, 2], latreg$beta[, 3])) | Statistical Modeling | https://osf.io/8fzns/ | 4H_PV_helper.R |
148 | Extract IRT likelihood of the second model (math, read and scie plus digital domains) | likeli_md <- IRT.likelihood(mod2) | Statistical Modeling | https://osf.io/8fzns/ | 4H_PV_helper.R |
149 | Compute latent regression of the latent ability on the conditioning variables (excl. the first column which is the student ID). But this time only the regression coefficients of the digital domains are computed freely. The rest is fixed at the values of the first model | latreg_md <- tam.latreg(likeli_md, Y = con_dat[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = "Ramsay"), beta.fixed = reg.coefs) | Statistical Modeling | https://osf.io/8fzns/ | 4H_PV_helper.R |
150 | Calculate the number of surveys completed per person | N <- as.data.frame(table(dat$id)) names(N) <- c("id", "N") dat <- merge(dat, N, by = "id", all.x = T) table(dat$N, useNA = "ifany") | Data Variable | https://osf.io/nxyh3/ | 01a_DataPrep_Study1.R |
151 | DOWNSAMPLING Loop to downsample data based on minimum time and distance between points Time parameter is set in "while(diff < time in minutes | dist > in meters)" downsampled pts that were less than 60min apart unless frog moved more than 20m during the time window | tracks_dsmpl <- data.frame() ids <- unique(tracks$id) for(i in ids) { traj = subset(tracks, tracks$id == i) for(i in 1:nrow(traj)) { diff <- difftime(traj$dt[i+1], traj$dt[i], units = "mins") delta_x <- traj$x_utm[i+1] - traj$x_utm[i] delta_y <- traj$y_utm[i+1] - traj$y_utm[i] dist <- sqrt(delta_x^2+delta_y^2) if(is.na(diff)) {break} while(diff <= 60 & dist < 20) { traj <- traj[-(i+1),] diff <- difftime(traj$dt[i+1], traj$dt[i], units = "mins") delta_x <- traj$x_utm[i+1] - traj$x_utm[i] delta_y <- traj$y_utm[i+1] - traj$y_utm[i] dist <- sqrt(delta_x^2+delta_y^2) if(is.na(diff)) {break} } } tracks_dsmpl <- bind_rows(tracks_dsmpl, traj) } tracks_dsmpl <- tracks_dsmpl %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = "min"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) tracks <- tracks_dsmpl | Data Variable | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
152 | Remove first row (tagging date) and last row (untagging day) per individual Sort the daily, group by id and add index of id+date as a new variable | daily %>% arrange(id) %>% group_by(id) %>% mutate(id_day = paste(id, date)) -> daily_grouped | Data Variable | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
153 | Join with summary stats | tracks_sum <- left_join(tracks_sum, duration) tracks_sum <- left_join(tracks_sum, relocs) tracks_sum <- left_join(tracks_sum, days_tracked) | Visualization | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
154 | Calculate average daily movement by behavioral category per individual for plotting | daily_beh_mean <- daily_beh %>% group_by(id, behavior) %>% dplyr::summarize(daily_dist = mean(daily_dist), max_dist = mean(max_dist), sex = first(sex), species = first(species)) | Visualization | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
155 | MCPs home range with 95% minimum convex ploygon (MCP95) | mcp95<- mcp(data_sp["id"], percent = 95, unout = "m2") | Data Variable | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
156 | 95% contour with hpi pluging method Define countour level | cont=c(95) | Visualization | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
157 | Initiate empty df to store the metadarta | allfrogs_info <- data.frame(ID = numeric(0), Cont = numeric(0), frog_id = character(0)) allfrogs_info <- data.frame(ID = numeric(0), Cont = numeric(0), frog_id = character(0)) | Data Variable | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
158 | BOXPLOTS: DAILY MOVEMENT FIGURE EXPORT: daily movement by sex | daily_plot <-tracks_sum%>% ggplot(aes(x= sex, y=log(mean_cumul), fill = species)) + theme_bw(20) + geom_boxplot(width= 0.6, outlier.shape = NA) + geom_jitter(aes(group = sex), position=position_jitterdodge(0.4), shape = 21, stroke = 1, color = "black", size = 2, alpha = 0.6) + scale_color_manual(values=c("black", "black")) + scale_fill_manual(values=c("#E7B800", "#0072B2","#FC4E07")) + theme(legend.position="none") + labs(y = "ln Daily movement (m)") + facet_wrap(~species, labeller = labeller(species=c("afemo" = "A. femoralis", "dtinc" = "D. tinctorius", "osylv" = "O. sylvatica"))) + theme(axis.title.x = element_blank(), axis.text.x = element_text(color = "black", size = 18), strip.text = element_text(face = "italic"), aspect.ratio = 4) + scale_x_discrete(labels= c("F", "M"), expand = expansion(add = 1)) daily_plot | Visualization | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
159 | Daily movement by sex and species, log transformed violin plots | ggplot(daily_select, aes(x= sex, y=log(daily_dist))) + theme_bw(20) + geom_violin(aes(fill = species)) + scale_fill_manual(values=c("#E7B800", "#0072B2","#FC4E07")) + theme(legend.position="none") + labs(y = "ln Daily movement (m)") + geom_jitter(aes(shape=sex, fill = species), position=position_jitterdodge(0.2), size = 4, alpha=0.2) + facet_wrap(~species, labeller = labeller( species=c(afemo ="A. femoralis", dtinc = "D. tinctorius", osylv = "O. sylvatica"))) + theme(axis.title.x = element_blank(), axis.text.x = element_text(color = "black", size = 18), strip.text = element_text(face = "italic")) + scale_x_discrete(labels= c("F", "M")) | Visualization | https://osf.io/3bpn6/ | all_spaceuse_dataproc.R |
160 | convert attributes from character to numeric | for(i in 1:3) { data[[paste0("opt",i,"cost")]] <- as.numeric(substring(data[[paste0("opt",i,"cost")]], 3, 7)) data[[paste0("opt",i,"sides")]] <- as.numeric(substring(data[[paste0("opt",i,"sides")]], 1, 2)) data[[paste0("opt",i,"deliveryTime")]] <- as.numeric(substring(data[[paste0("opt",i,"deliveryTime")]], 2, 3)) } | Data Variable | https://osf.io/tbczv/ | exp1aDead-MNL-SAT-A.r |
161 | Step 9: Create time difference variable, and the final dependent variable (media_success) calculating the time difference between each documentpair | result$date_diff <- as.Date(as.character(result$news_date), format="%Y-%m-%d")- as.Date(as.character(result$date), format="%Y-%m-%d") | Data Variable | https://osf.io/hfy4k/ | prep.analysis.data.R |
162 | Remove trials with RT < 4710 ms (i.e., presses before disambiguating information in sentence) | dataset_exp1 <- dataset_exp1[which(dataset_exp1$rt > 4710), ] | Data Variable | https://osf.io/37rfb/ | prediction_analyses.R |
163 | Add 'trackloss' column (if not looking at IA_1 or IA_2, then trackloss 1) | dataset_exp1 <- within(dataset_exp1, { trackloss <- ifelse(average_target_sample_count_proportion == 0 & average_distractor_sample_count_proportion == 0, 1, 0) }) | Data Variable | https://osf.io/37rfb/ | prediction_analyses.R |
164 | Ttests (using participantaveraged data) | num_sub_exp1 <- length(unique((eyetrackingr.data.exp1$participant_number))) threshold_t_exp1 <- qt(p = 1 - .05 / 2, df = num_sub_exp1 - 1) # Pick threshold for t based on alpha = .05, two-tailed df_timeclust_exp1 <- make_time_cluster_data(response_time_exp1, test = "t.test", paired = TRUE, predictor_column = "trial_condition_new", threshold = threshold_t_exp1 ) | Statistical Test | https://osf.io/37rfb/ | prediction_analyses.R |
165 | MSE for each participant | MSE.Sys.VAR.i = lapply(1:fold, function(k) MSE.k[[k]]$MSE.ki) | Data Variable | https://osf.io/rs6un/ | MSE.VAR.Sys.R |
166 | use mutate and the dffit fit information to make the data for mal | mal_fit <- fit_num %>% mutate(fit = (s)/(1.196506 + s)) dex_fit <- fit_num %>% mutate(fit = (s)/(62.736799 + s)) lac_fit <- fit_num %>% mutate(fit = (s)/(43.839063 + s)) maldex_fit <- fit_num %>% mutate(fit = (s)/(7.253739 + s)) | Statistical Modeling | https://osf.io/9e3cu/ | titration_answers.R |
167 | Logistic Mixed Effects Regression for Original Critical Targets | log_reg_data <- cards_long %>% filter(card == "card_1" | card == "card_3" | card == "card_6" | card == "card_7" | card == "card_9" | card == "card_14" ) %>% mutate( distance = case_when( card == "card_1" | card == "card_3" | card == "card_6" ~ 0, card == "card_7" | card == "card_9" | card == "card_14" ~ 1 ), comf_acquaint = scale(comf_acquaint, scale = FALSE), comf_close = scale(comf_close, scale = FALSE), approp_acquaint = scale(approp_acquaint, scale = FALSE), approp_close = scale(approp_close, scale = FALSE) ) log_reg_data$amit_wording <- factor(log_reg_data$amit_wording, levels = c("original", "new")) h1_melogreg <- glmer(choice ~ distance + amit_wording + (1 + distance|id) + (1|card), data = log_reg_data, family = binomial(link = "logit")) h1_melogreg_inter <- glmer(choice ~ distance * amit_wording + (1 + distance|id) + (1|card), data = log_reg_data, family = binomial(link = "logit")) h1_model_comparison <- anova(h1_melogreg, h1_melogreg_inter) | Statistical Modeling | https://osf.io/bkuwa/ | main_analysis_code.R |
168 | Highest density interval ' ' This is a function that will calculate the highest density interval from a ' posterior sample. ' ' The default is to calcualte the highest 95 percent interval. It can be used ' with any numeric vector instead of having to use one of the specific MCMC ' classes. This function has been adapted from John K. Kruschke (2011). Doing ' Bayesian Data Analaysis: A Tutorial with R and BUGS. ' ' @param x Numeric vector of a distribution of data, typically a posterior ' sample ' @param prob Width of the interval from some distribution. Defaults to `0.95`. ' @param warn Option to turn off multiple sample warning message Must be in the ' range of `[0,1]`. ' @return Numeric range ' @export ' @examples ' x < qnorm(seq(1e04, .9999, length.out1001)) ' hdi_95 < hdi(x, .95) ' hdi_50 < hdi(x, .50) ' ' hist(x, br50) ' abline(vhdi_95, col"red") ' abline(vhdi_50, col"green") ' ' x < exp(seq(pi * (1 (1/16)), pi, len 1000)) ' x < c(x, rev(x)[1]) ' x < c(x, x) ' plot(sort(x), type"l") ' plot(density(x, adjust0.25)) ' abline(vhdi(x, p.49), col2) ' abline(vhdi(x, p.50), col3) | hdi <- function(x, prob=0.95, warn=TRUE) { if (anyNA(x)) { stop("HDI: ", "x must not contain any NA values.", call.=FALSE) } N <- length(x) if (N < 3) { if (warn) { warning("HDI: ", "length of `x` < 3.", " Returning NAs", call.=FALSE) } return(c(NA_integer_, NA_integer_)) } x_sort <- sort(x) window_size <- as.integer(floor(prob * length(x_sort))) if (window_size < 2) { if (warn) { warning("HDI: ", "window_size < 2.", " `prob` is too small or x does not ", "contain enough data points.", " Returning NAs.", call.=FALSE) } return(c(NA_integer_, NA_integer_)) } lower <- seq_len(N - window_size) upper <- window_size + lower | Statistical Modeling | https://osf.io/nd9yr/ | ordinal_helper_functions.R |
169 | test whether Condition predicts exclusion probability | Cond_exclude <- table(data$Cond, data$exclude) chisq.test(Cond_exclude) rm(Cond_exclude) | Data Variable | https://osf.io/sb3kw/ | Study2B_analyses.R |
170 | Fix the following script by: Take our dataset (df) , and then group it by our explanatory variable (Condition), and then Summarise this dataset by Creating a new variable in this summary dataset called Mean Mean takes the mean of our Response variable, and ignores NA. Repeat for standard deviation. We've added on one more column in our summary dataset N. N is the number of observations per group. To do so, we want to add up How many times our DV is not missing (per group). When you think it is a complete, save the script. The name of the script above will go from red to black. That is how you know you've saved it. You can use ctrl+S to save, or click the floppy disk above. We want to know how the means and standard deviations and Ns of Hire_Rating differed. | sum_hire <- | Data Variable | https://osf.io/9vr6q/ | summarise2.R |
171 | Calculate R squared for each model | r.squaredGLMM(m1) #indegree r.squaredGLMM(m2) #outdegree r.squaredGLMM(m3) #betweenness r.squaredGLMM(m4) #outcloseness r.squaredGLMM(m5) #incloseness r.squaredGLMM(m6) #local clustering r.squaredGLMM(m7) #average shortest path length r.squaredGLMM(m8) #eigenvector r.squaredGLMM(m9) #outstrength r.squaredGLMM(m10) #instrength | Statistical Test | https://osf.io/wc3nq/ | 6) agr_summer_model.R |
172 | split the data into two halves | d$ItemNo = parse_number(as.character(d$ItemNo)) de = d %>% filter(ItemNo %% 2 == 0) do = d %>% filter(ItemNo %% 2 == 1) | Data Variable | https://osf.io/cd5r8/ | R_Code_Buffinton_MorganShort.R |
173 | Zscore transform RTs, compute zRT difference, remove NAs for the 2 groups | de <- de |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) do <- do |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) de <- de |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) | Data Variable | https://osf.io/cd5r8/ | R_Code_Buffinton_MorganShort.R |
174 | fit models for even and odd numbered items (Bayesian & Generalized LMM) | m_de = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = de, control=lmerControl(optimizer = "nloptwrap", optCtrl = list(algorithm = "NLOPT_LN_NELDERMEAD", maxit = 2e5))) m_do = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = do, control=lmerControl(optimizer = "nloptwrap", optCtrl = list(algorithm = "NLOPT_LN_NELDERMEAD", maxit = 2e5))) m_de = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = de, control=lmerControl(optimizer = "nloptwrap", optCtrl = list(algorithm = "NLOPT_LN_NELDERMEAD", maxit = 2e5))) m_do = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = do, control=lmerControl(optimizer = "nloptwrap", optCtrl = list(algorithm = "NLOPT_LN_NELDERMEAD", maxit = 2e5))) | Statistical Modeling | https://osf.io/cd5r8/ | R_Code_Buffinton_MorganShort.R |
175 | fit lmm to the whole dataset | m <- blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = d_trimmed, control=lmerControl(optimizer = "nloptwrap", optCtrl = list(algorithm = "NLOPT_LN_NELDERMEAD", maxit = 2e5))) summary(m) | Statistical Modeling | https://osf.io/cd5r8/ | R_Code_Buffinton_MorganShort.R |
176 | calculate shared and unique variance | x1 = rsq - rx2y^2 x2 = rsq - rx1y^2 x1_x2 = rsq - x1 - x2 c(x1, x2, x1_x2, rsq) } semipartial_corr = function(r12, r13, r23) { | Statistical Modeling | https://osf.io/sqfnt/ | Goal |
177 | create empty df | df <- data.frame( `rx1y` = numeric(), `rx2y` = numeric(), `rx1x2` = numeric(), `rx1y_x2` = numeric(), `rx2y_x1` = numeric(), `x1` = numeric(), `x2` = numeric(), `Common` = numeric(), `Total` = numeric() ) | Data Variable | https://osf.io/sqfnt/ | Goal |
178 | > Left panel: QQplot (uniform distribution) > Right panel: Residuals against predicted values;; shaded (due to sample size) with extreme residuals colored red, and 3) MAIN EFFECTS OF KEY VARIABLES AspElevForm | emmip(PrefRate, ~AspElevForm, type = "response", CIs = TRUE) (emm <- emmeans(PrefRate, specs = ~AspElevForm, type = "response")) pairs(emm) | Visualization | https://osf.io/2sz48/ | Model_Preference.R |
179 | Operator on x, rescale to be in y unit | Hlist <- list() for(i in seq(nt)){ Hlist[[paste0("H",i)]] <- with(data, t(apply(data[[i]]$sdfp, 1, function(x) x*data[[i]]$prior)) %*% A) } H<-bdiag(unlist(Hlist)) strl <- function(x){ x[x < max(x)/1000] <- 0 return(x) } H <- as(t(apply(H, 1, strl)), 'dgCMatrix') #Make sparser | Data Variable | https://osf.io/53w96/ | INLA_GHG_GMD.R |
180 | Creates the variable 'distances', a 540 X 3 distance matrix between stimuli and category prototypes. | distances = as.matrix(pdist(coordinates, prototypes)) | Data Variable | https://osf.io/hrf5t/ | runPrototype.R |
181 | calculate similarity coefficient (mean of signif_line$similarity.scores) | mean(signif_line$similarity.scores[,2]) mean(coef_line$similarity.scores[,2]) mean(varimp_line$similarity.scores[,2]) | Statistical Modeling | https://osf.io/3gfqn/ | VADIS_genitives_outer_circle.R |
182 | Calculate Priestley Taylor Ep | data$PT.Ep <- 1.26/lambda.w*deltav/(deltav+rho.air)*Rn | Data Variable | https://osf.io/5ezfk/ | ETfun.R |
183 | plot hit, FA, miss, corrRejection for each effect size/sample size combo | ap <- aggregate(hits ~ criterion+sampleSize+effectSizes, data=allDt,mean) ap2 <- aggregate(fa ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(ap,ap2,by=c("criterion","sampleSize","effectSizes")) ap2 <- aggregate(miss ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(hFA,ap2,by=c("criterion","sampleSize","effectSizes")) ap2 <- aggregate(corrRej ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(hFA,ap2,by=c("criterion","sampleSize","effectSizes")) hFA$hits <- hFA$hits / numStudies hFA$fa <- hFA$fa / numStudies hFA$corrRej <- hFA$corrRej / numStudies hFA$miss <- hFA$miss / numStudies | Visualization | https://osf.io/hzncs/ | Witt_SDT_Simulations_SeveralTests.R |
184 | plot errors only (FA and misses, separately) Plot ROC dist | aa <- aggregate(ROCdist ~ criterion, data=allDt,mean) ab <- aggregate(ROCdist ~ criterion, data=allDt,sd) ab$ci <- qnorm(.975) * ab$ROCdist / sqrt(numStudies) titleText <- ifelse(length(sSizesAll) < 2, paste("N=",sSizesAll[1],"d=",effSzAll[1]), paste("firstES:",effSzAll[1])) plot(seq(1:length(crits)),aa$ROCdist,bty="l",xaxt="n",xlab="Criterion for Statistical Significance",ylab="Distance to Perfection",pch=19,col=rainbow(length(crits)),cex=2,ylim=c(0,max(aa$ROCdist)+max(ab$ci)),main = titleText) axis(side=1,at=seq(1:length(crits)),labels = crits) for(i in 1:length(crits)) { segments(i,aa$ROCdist[i] - ab$ci[i],i,aa$ROCdist[i]+ab$ci[i]) } | Visualization | https://osf.io/hzncs/ | Witt_SDT_Simulations_SeveralTests.R |
185 | calculate the variance and bias for the log(DR) | log.var <- ((res$se.tau2)^2)/4*1/(sum(w.star))^2*(sum(1/(res$vi + res$tau2)^2))^2 log.sd <- sqrt(log.var) bias <- 1/2*(res$se.tau2)^2*(1/2/sum(w.star)^2 - 1/sum(w.star)*sum(1/(res$vi + res$tau2)^3)) | Statistical Modeling | https://osf.io/gwn4y/ | Hospital_Stay_of_Stroke_Patients_forest_plot.R |
186 | create the forest plot including extra information (study names, weights, observed effects and associated variances). | f <- forest(res, col = "blue", border = "blue", ylim = c(-8,12), xlim = c(-4,5), pch = 19, slab = dat$study, showweights = FALSE, addfit = FALSE, refline = FALSE, ilab = cbind(paste(format(round(res$yi,2), nsmall = 2)) , paste(format(round(res$vi, 2), nsmall = 2)), paste(format(round(res$ni, 2))), paste(format(weights, nsmall = 2)), paste(format(weights.f, nsmall = 2))) , ilab.xpos = c(-3.15,-2.75, -2.35,-1.5,-0.75), ilab.pos = 2, efac = 0, digits = 2) | Visualization | https://osf.io/gwn4y/ | Hospital_Stay_of_Stroke_Patients_forest_plot.R |
187 | STEP 2: Filter the data in overall.data table according to preregistered inclusion criteria replace 0 with NA in outcome column | overall.data$LTScreenOut[overall.data$LTScreenOut == 0] <- NA #replace NAs with 0 for averaging in the next steps overall.data$LTObjectOut[overall.data$LTObjectOut == 0] <- NA #replace NAs with 0 for averaging in the next steps overall.data$FirstLookDurationObjectOut[overall.data$FirstLookDurationObjectOut == 0] <- NA #replace NAs with 0 for averaging in the next steps | Data Variable | https://osf.io/mp9td/ | TablePrep_Third.R |
188 | fit indices function for the SEM and FA analyses | fa.CFI<-function(x){ nombre<-paste(x,"CFI",sep = ".") nombre<- ((x$null.chisq-x$null.dof)-(x$STATISTIC-x$dof))/(x$null.chisq-x$null.dof) return(nombre) } | Statistical Modeling | https://osf.io/bhrwx/ | evaluation_data_analysis.R |
189 | check Descriptive statistics gender breakdown | df <- eval_data %>% group_by(gender) %>% summarise(counts = n()) df | Statistical Test | https://osf.io/bhrwx/ | evaluation_data_analysis.R |
190 | Decompose variance Estimate multilevel model (without predictors) | m_logistic <- lmer(estimate ~ 1 + (1|controls) + (1|age) + (1|year) + (1|age:year) + (1|age:controls) + (1|year:controls), data = results %>% rename(age = age_group)) | Statistical Modeling | https://osf.io/m72gb/ | analysis_privacysetting.r |
191 | Visualize four first principal components | fig.df <- data.frame(HR = c(FPCAdense$phi[,1], FPCAdense$phi[,2], FPCAdense$phi[,3], FPCAdense$phi[,4]), pc = c(rep(1, 51), rep(2, 51), rep(3, 51), rep(4, 51)), Time= rep(FPCAdense$workGrid, 4)) all.four <- ggplot(fig.df, aes(x=Time, y=HR))+geom_line()+ geom_vline(xintercept=.20, linetype="dashed")+ facet_grid(.~pc)+ theme_bw() all.four ggsave("../figures/figure5.png", width=6, height=4) | Visualization | https://osf.io/qj86m/ | 8_fda_socaccount.R |
192 | SPATIOTEMPORAL VARIABLES AND HOMING SUCCESS Join territory center pts to the dataframe to calculate translocation distance and homing success | trajectory <- trajectory.df %>% left_join(territory_centers, by = "id")%>% mutate(trans_dist = sqrt((x_home - first(x_utm))^2+(y_home - first(y_utm))^2), trans_group = ifelse(trans_dist > 100, "200m", "50m"))%>% dplyr::select(id, sex, trans_group, trans_dist, dt, x_utm, y_utm, x_home, y_home, time_lag_min = time_diff, dist) | Data Variable | https://osf.io/3bpn6/ | dt_homing_dataproc.R |
193 | Tally the number relocations for each individual per day | trajectory.df %>% group_by(id, date) %>% count(id) -> relocs.perday | Data Variable | https://osf.io/3bpn6/ | dt_homing_dataproc.R |
194 | Plot full circular info by sex start with blank plot, then add groupspecific points | dev.new();; par(mai = c(1, 1, 0.1,0.1)) par(mar=c(0.5, 0.5, 0.5, 0.5)) pdf("dt_circ_50m.pdf") plot(corient50_m, bg = rgb(0, 0.749, 0.769), pch = 21, cex = 1.5, lwd = 2, stack = TRUE, bin = 60, xlim = c(-1.2, 1.2), ylim = c(-1.2, 1.2), sep = 0.05, shrink = 1, tcl.text = -0.125, control.circle=circle.control(lwd = 2)) ticks.circular(circular(seq(0,2*pi,pi/2)), tcl=0.075) par(new = T) plot(corient50_f, bg=rgb(0.973, 0.463, 0.427), pch = 21, cex=1.5, lwd = 2, stack=T, bins=60, sep = -0.05, shrink= 1.3, axes = FALSE, control.circle=circle.control(lwd = 1)) arrows.circular(mean(corient50_m), y = rho.circular(corient50_m), col = rgb(0, 0.749, 0.769), lwd = 5) arrows.circular(mean(corient50_f), y = rho.circular(corient50_f), col = rgb(0.973, 0.463, 0.427), lwd = 5) par(new = T) plot(corient50_m, col = NA, shrink= 2.5, axes = FALSE, control.circle=circle.control(lty = 2, lwd = 1)) ticks.circular(circular(seq(0,2*pi,pi/8)), tcl=0.2) lines(density.circular(corient50_m, bw=30), shrink= 1, col = rgb(0, 0.749, 0.769, 0.7), lwd=2, lty=1) lines(density.circular(corient50_f, bw=30), col = rgb(0.973, 0.463, 0.427, 0.7), lwd=2, lty=1) dev.new();; par(mai = c(1, 1, 0.1,0.1)) par(mar=c(0.5, 0.5, 0.5, 0.5)) pdf("dt_circ_200m.pdf") plot(corient200_m, bg = rgb(0, 0.749, 0.769), pch = 21, cex = 1.5, lwd = 2, stack = TRUE, bin = 60, xlim = c(-1.2, 1.2), ylim = c(-1.2, 1.2), sep = 0.05, shrink = 1, tcl.text = -0.125, control.circle=circle.control(lwd = 2)) ticks.circular(circular(seq(0,2*pi,pi/2)), tcl=0.075) par(new = T) plot(corient200_f, bg=rgb(0.973, 0.463, 0.427), pch = 21, cex=1.5, lwd = 2, stack=T, bins=60, sep = -0.05, shrink= 1.3, axes = FALSE, control.circle=circle.control(lwd = 1)) arrows.circular(mean(corient200_m), y = rho.circular(corient200_m), col = rgb(0, 0.749, 0.769), lwd = 5) arrows.circular(mean(corient200_f), y = rho.circular(corient200_f), col = rgb(0.973, 0.463, 0.427), lwd = 5) par(new = T) plot(corient200_m, col = NA, shrink= 2.5, axes = FALSE, control.circle=circle.control(lty = 2, lwd = 1)) ticks.circular(circular(seq(0,2*pi,pi/8)), tcl=0.2) lines(density.circular(corient200_m, bw=30), shrink= 1, col = rgb(0, 0.749, 0.769, 0.7), lwd=2, lty=1) lines(density.circular(corient200_f, bw=30), col = rgb(0.973, 0.463, 0.427, 0.7), lwd=2, lty=1) | Visualization | https://osf.io/3bpn6/ | dt_homing_dataproc.R |
195 | Helper function to prepare raw keyboard data ' ' @author F. Bemmann ' @family Preprocessing function ' @description this function unfolds json data into one column per keyvalue pair ' @export | parseJsonColumnSensing = function(df, column_name){ parseJsonColumn = function(x){ str_c("[ ", str_c(x, collapse = ",", sep=" "), " ]") %>% jsonlite::fromJSON(flatten = T) %>% as_tibble() } df2 = df %>% select(user_uuid,client_event_id,!!column_name) %>% filter(!is.na(!!rlang::sym(column_name))) %>% map_dfc(.f = parseJsonColumn) %>% distinct() colnames(df2)[1:2] = c("user_uuid", "client_event_id") df = left_join(df, df2, by = c("user_uuid", "client_event_id")) df[,column_name] = NULL return(df) } | Data Variable | https://osf.io/b7krz/ | helper_JsonFormat.R |
196 | Separate ANOVAs for each subscale Pairwise ttests with adjusted alpha level of .05/4 .0125 (because there are four relevant comparisons) | summary(aov(value~Fan*time+Error(VP_t0/(time)),data=dat[dat$scale=="Attentiveness",])) pairwise.t.test(dat[dat$scale=="Attentiveness",]$value,paste(dat[dat$scale=="Attentiveness",]$Fan,dat[dat$scale=="Attentiveness",]$time),p.adj="none") | Statistical Test | https://osf.io/xcthg/ | 0-Script.R |
197 | Models .. Contrasts (Reverse) Helmert for cond speech, cond mask Treatment for speech_mask Sum for group Factor variables | df <- df %>% mutate( group = factor(group), # oc, pd gender = factor(intake_gender), # f, m cond_mask = factor(cond_mask, levels = c("nm","sm","kn")), cond_speech = factor(cond_speech, levels = c("habitual","clear","loud")) ) %>% mutate(speech_mask = paste(cond_speech,cond_mask,sep="_")) levels(df$group) # oc, pd levels(df$gender) # f, m levels(df$cond_speech) # habitual, clear, loud levels(df$cond_mask) # nm, sm, kn | Statistical Modeling | https://osf.io/5s34w/ | simpd_helper.R |
198 | 4) Descriptives and data inspection total duration on survey | stat.desc(working_file %>% group_by(pp) %>% slice(1) %>% pull(duration)) | Data Variable | https://osf.io/g8kbu/ | dataAnalysisRewardAppsSurvey.R |
199 | bayesian ttests | ttestBF(wide$high, wide$neutral, paired = TRUE) ttestBF(wide$low, wide$neutral, paired = TRUE) ttestBF(wide$high, wide$low, paired = TRUE) formatC(3057348376, format = "e", digits = 2) | Statistical Test | https://osf.io/g8kbu/ | dataAnalysisRewardAppsSurvey.R |
200 | Confidence interval as a vector | result <- c("lower" = vec_mean - error, "upper" = vec_mean + error) return(result) } | Statistical Modeling | https://osf.io/92e6c/ | fill_summary_table.R |
Subsets and Splits