I have fit a robust random-effects meta-regression model using metafor package in R.
My full data, as well as reproducible R code appear below.
Questions:
(1) What are the meaning and interpretation of grey-colored diamonds appearing over CIs?
(2) I won't get an overall mean effect when I have moderators, correct?
library(metafor)
d <- read.csv("https://raw.githubusercontent.com/izeh/m/master/d.csv", h = T) ## DATA
res <- robust(rma.uni(yi = dint, sei = SD, mods = ~es.type, data = d, slab = d$study.name),
cluster = d$id)
forest(res)
1) Quoting from help(forest.rma): "For models involving moderators, the fitted value for each study is added as a polygon to the plot." So, the grey-colored diamonds (or polygons) are the fitted values and the width of the diamonds/polygons reflects the width of the CI for the fitted values.
2) No, since there is no longer a single overall effect when your model includes moderators.
Related
I ran a conditional random forest regression model using the cforest function from the party package because I have both categorical and continuous predictor variables that are correlated with each other, and a continuous outcome variable.
Here is my code to run the conditional random forest model, obtain out-of-bag estimates, and estimate the permutation variable importance.
# 1. fit the random forest
crf <- party::cforest(Y ~ ., data = df,
controls = party::cforest_unbiased(ntree = 10000, mtry = 7))
# 2. obtain out-of-bag estimates
pred_oob <- as.data.frame(caret::predict(crf, OOB = T, newdata = NULL))
# 3. estimate permutation variable importance
vi <- permimp::permimp(crf, condition = T, threshold = 0.5, nperm = 1000, OOB = T,
mincriterion = 0)))
I would like to visualize the minimal depth distribution and calculate mean minimal depth similar to the output from the RandomForestExplainer package. However, the RandomForestExplainer package only takes in objects from the randomForest function in the randomForest package. It's not an option for me to use this function due to the nature of my data (described above).
I have been combing the internet and have not been able to find a solution. Can someone point me to a way to visualize the minimal depth distribution for all predictors and calculate the mean minimal depth?
I am doing a multi-level meta-analysis. Many studies have several subgroups. When I make a forest plot studies are presented as subgroups. There are 60 of them, however, I would like to plot studies according to the study, then it would be 25 studies and it would be more appropriate. Does anyone have an idea how to do this forest plot?
I did it this way:
full.model <- rma.mv(yi = yi,
V = vi,
slab = Author,
data = df,
random = ~ 1 | Author/Study,
test = "t",
method = "REML")
forest(full.model)
It is not clear to me if you want to aggregate to the Author level or to the Study level. If there are multiple rows of data for particular studies, then the model isn't really complete and you would want to add another random intercept for the level of the estimates within studies. Essentially, the lowest random effect should have as many values for nlvls in the output as there are estimates (k).
Let's first tackle the case where we have a multilevel structure with two levels, studies and multiple estimates within studies (for some technical reasons, some might call this a three-level model, but let's not get into this). I will use a fully reproducible example for illustration purposes, using the dat.konstantopoulos2011 dataset, where we have districts and schools within districts. We fit a multilevel model of the type as you have with:
library(metafor)
dat <- dat.konstantopoulos2011
res <- rma.mv(yi, vi, random = ~ 1 | district/school, data=dat)
res
We can aggregate the estimates to the district level using the aggregate() function, specifying the marginal var-cov matrix of the estimates from the model to account for their non-independence (note that this makes use of aggregate.escalc() which only works with escalc objects, so if it is not, you need to convert the dataset to one - see help(aggregate.escalc) for details):
agg <- aggregate(dat, cluster=dat$district, V=vcov(res, type="obs"))
agg
You will find that if you then fit an equal-effects model to these estimates based on the aggregated data that the results are identical to what you obtained from the multilevel model (we use an equal-effects model since the heterogeneity accounted for by the multilevel model is already encapsulated in vcov(res, type="obs")):
rma(yi, vi, method="EE", data=agg)
So, we can now use these aggregated values in a forest plot:
with(agg, forest(yi, vi, slab=district))
My guess based on your description is that you actually have an additional level that you should include in the model and that you want to aggregate to the intermediate level. This is a tad more complicated, since aggregate() isn't meant for that. Just for illustration purposes, say we use year as another (higher) level and I will mess a bit with the data so that all three variance components are non-zero (again, just for illustration purposes):
dat$yi[dat$year == 1976] <- dat$yi[dat$year == 1976] + 0.8
res <- rma.mv(yi, vi, random = ~ 1 | year/district/school, data=dat)
res
Now instead of aggregate(), we can accomplish the same thing by using a multivariate model, including the intermediate level as a factor and using again vcov(res, type="obs") as the var-cov matrix:
agg <- rma.mv(yi, V=vcov(res, type="obs"), mods = ~ 0 + factor(district), data=dat)
agg
Now the model coefficients of this model are the aggregated values and the var-cov matrix of the model coefficients is the var-cov matrix of these aggregated values:
coef(agg)
vcov(agg)
They are not all independent (since we haven't aggregated to the highest level), so if we want to check that we can obtain the same results as from the multilevel model, we must account for this dependency:
rma.mv(coef(agg), V=vcov(agg), method="EE")
Again, exactly the same results. So now we use these coefficients and the diagonal from vcov(agg) as their sampling variances in the forest plot:
forest(coef(agg), diag(vcov(agg)), slab=names(coef(agg)))
The forest plot cannot indicate the dependency that still remains in these values, so if one were to meta-analyze these aggregated values using only diag(vcov(agg)) as their sampling variances, the results would not be identical to what you get from the full multilevel model. But there isn't really a way around that and the plot is just a visualization of the aggregated estimates and the CIs shown are correct.
You need to specify your own grouping in a new column of data and use this as the new random effect:
df$study_group <- c(1,1,1,2,2,3,4,5,5,5) # example
full.model <- rma.mv(yi = yi,
V = vi,
slab = Author,
data = df,
random = ~ 1 | study_group,
test = "t",
method = "REML")
forest(full.model)
I have build a binary logistic regression for churn prediction in Rstudio. Due to the unbalanced data used for this model, I also included weights. Then I tried to find the optimum cutoff by try and error, however To complete my research I have to incorporate ROC curves to find the optimum cutoff. Below I provided the script I used to build the model (fit2). The weight is stored in 'W'. This states that the costs of wrongly identifying a churner is 14 times as large as the costs of wrongly identifying a non-churner.
#CH1 logistic regression
library(caret)
W = 14
lvl = levels(trainingset$CH1)
print(lvl)
#if positive we give it the defined weight, otherwise set it to 1
fit_wts = ifelse(trainingset$CH1==lvl[2],W,1)
fit2 = glm(CH1 ~ RET + ORD + LVB + REVA + OPEN + REV2KF + CAL + PSIZEF + COM_P_C + PEN + SHOP, data = trainingset, weight=fit_wts, family=binomial(link='logit'))
# we test it on the test set
predlog1 = ifelse(predict(fit2,testset,type="response")>0.5,lvl[2],lvl[1])
predlog1 = factor(predlog1,levels=lvl)
predlog1
confusionMatrix(pred,testset$CH1,positive=lvl[2])
For this research I have also build ROC curves for decision trees using the pROC package. However, of course the same script does not work the same for a logistic regression. I have created a ROC curve for the logistic regression using the script below.
prob=predict(fit2, testset, type=c("response"))
testset$prob=prob
library(pROC)
g <- roc(CH1 ~ prob, data = testset, )
g
plot(g)
Which resulted in the ROC curve below.
How do I get the optimum cut off from this ROC curve?
Getting the "optimal" cutoff is totally independent of the type of model, so you can get it like you would for any other type of model with pROC. With the coords function:
coords(g, "best", transpose = FALSE)
Or directly on a plot:
plot(g, print.thres=TRUE)
Now the above simply maximizes the sum of sensitivity and specificity. This is often too simplistic and you probably need a clear definition of "optimal" that is adapted to your use case. That's mostly beyond the scope of this question, but as a starting point you should a look at Best Thresholds section of the documentation of the coords function for some basic options.
In the past, I had used the sjp.glmer from the package sjPlot to visualize the different slopes from a generalized mixed effects model. However, with the new package, I can't figure out how to plot the individual slopes, as in the figure for the probabilities of fixed effects by (random) group level, located here
Here is the code that, I think, should allow for the production of the figure. I just can't seem to get it in the new version of sjPlot.
library(lme4)
library(sjPlot)
data(efc)
# create binary response
efc$hi_qol = 0
efc$hi_qol[efc$quol_5 > mean(efc$quol_5,na.rm=T)] = 1
# prepare group variable
efc$grp = as.factor(efc$e15relat)
# data frame for 2nd fitted model
mydf <- na.omit(data.frame(hi_qol = as.factor(efc$hi_qol),
sex = as.factor(efc$c161sex),
c12hour = as.numeric(efc$c12hour),
neg_c_7 = as.numeric(efc$neg_c_7),
grp = efc$grp))
# fit 2nd model
fit2 <- glmer(hi_qol ~ sex + c12hour + neg_c_7 + (1|grp),
data = mydf,
family = binomial("logit"))
I have tried to graph the model using the following code.
plot_model(fit2,type="re")
plot_model(fit2,type="prob")
plot_model(fit2,type="eff")
I think that I may be missing a flag, but after reading through the documentation, I can't find out what that flag may be.
Looks like this might do what you want:
(pp <- plot_model(fit2,type="pred",
terms=c("c12hour","grp"),pred.type="re"))
type="pred": plot predicted values
terms=c("c12hour", "grp"): include c12hour (as the x-axis variable) and grp in the predictions
pred.type="re": random effects
I haven't been able to get confidence-interval ribbons yet (tried ci.lvl=0.9, but no luck ...)
pp+facet_wrap(~group) comes closer to the plot shown in the linked blog post (each random-effects level gets its own facet ...)
Ben already posted the correct answer. sjPlot uses the ggeffects-package for marginal effects plot, so an alternative would be using ggeffects directly:
ggpredict(fit2, terms = c("c12hour", "grp"), type="re") %>% plot()
There's a new vignette describing how to get marginal effects for mixed models / random effects. However, confidence intervals are currently not available for this plot-type.
The type = "ri.prob" option in the linked blog-post did not adjust for covariates, that's why I first removed that option and later re-implemented it (correctly) in ggeffects / sjPlot. The confidence intervals shown in the linked blog-post are not correct, either. Once I figure out a way how to obtain CI or prediction intervals, I'll add this option as well.
I'm trying to plot ROC curve of a random forest classification. Plotting works, but I think I'm plotting the wrong data since the resulting plot only has one point (the accuracy).
This is the code I use:
set.seed(55)
data.controls <- cforest_unbiased(ntree=100, mtry=3)
data.rf <- cforest(type ~ ., data = dataset ,controls=data.controls)
pred <- predict(data.rf, type="response")
preds <- prediction(as.numeric(pred), dataset$type)
perf <- performance(preds,"tpr","fpr")
performance(preds,"auc")#y.values
confusionMatrix(pred, dataset$type)
plot(perf,col='red',lwd=3)
abline(a=0,b=1,lwd=2,lty=2,col="gray")
To plot a receiver operating curve you need to hand over continuous output of the classifier, e.g. posterior probabilities. That is, you need to predict (data.rf, newdata, type = "prob").
predicting with type = "response" already gives you the "hardened" factor as output. Thus, your working point is implicitly fixed already. With respect to that, your plot is correct.
side note: in bag prediction of random forests will be highly overoptimistic!