I am new to R and having some trouble with plotting svm models.
1)How can we plot and analyze mulit variable SVM regression model results.
library(e1071)
set.seed(3)
data = data.frame(matrix(rnorm(100*5), nrow=100))
train=data[1:70,]
test=data[71:100,]
fit = svm(X1 ~ ., data=train)
summary(fit)
pred=predict(fit,test)
2) Assume one of the variable (eg: X2) contains qualitative data (eg: high,low and medium) instead of quantitative data, then how should we plot
In short: you cannot. There is no way to visualize an object that is more than 3-dimensional.
What you can do is to deal with some simplification, approximation, etc. you often visualize characteristic of the model and not the model itself. For example one might plot:
relation between error metric (like R2) vs. some hyperparameter (regularization strength, kernel width, size of the training sets etc.)
find two most significant dimensions of the dataset and plot your model as 3d surface on top of these two dimensions only
if your dimensionality is not very high you can do pairplots, so visualize each pair of dimensions -> as it requires d(d-1)/2 plots, thus for d=5 it is just 10 plots.
many other characteristic important from the perspective of your experiment
Related
I would like to fit a single model to several independent datasets in R using the spatstat package. Here, I have 3 independent datasets (ppp objects called NMJ1, NMJ2, and NMJ3), to which I want to fit a common model. The way to go should be to use the mppm function:
data <- listof(NMJ1,NMJ2,NMJ3)
data <- hyperframe(X=1:3, Points=data)
r <- matrix(c(120, 240, 240, 90), nrow = 2, ncol = 2)
model <- mppm(Points ~marks*abs(sqrt(x^2+y^2)), data, MultiStrauss(r))
In the model I am fitting, the intensity is a function of the distance to the center of the window and I supposed a MultiStrauss interaction scheme.
However, the mppm function is going to fit each dataset independently. When typing subfits(model), the fitted trend coefficients are the same for each dataset, but not the gamma coefficients. Similarly, in plotting the results of simulate(model), I observe significant and consistent differences between the 3 plots.
What is the best way to handle independent datasets (repetition of samples from the same model) in spatstat ?
This is a bug.
Your code is correct for this purpose. (Namely, when we want the same interaction coefficients to apply to all of the point patterns.)
There is a bug in the function subfits in the package spatstat.core.
The fitted model returned by mppm is correct, but the list of sub-models returned by subfits is partially incorrect.
The bug will be fixed shortly, in the development version spatstat.core 2.3-0.011 available from the GitHub repository
How can I plot predicted survival curves of a continuous covariate (let's say 20th and 80th percentile of the value) using the corrected group prognosis method as implemented in R by Therneau
For example,
library(survival)
library(survminer)
fit <- coxph( Surv(stop, event) ~ size + strata(rx), data = bladder )
ggadjustedcurves(fit, data=bladder, method = "conditional", strata=rx)
Now, this is useful because I am given two survival curves that are stratified by rx (either 0 or 1) and the conditional method is being acted upon the bladder data set. However, let's say I would like to use the marginal method but not stratify and instead plot my continuous covariate at 20th and 80th value but also re-balance the subpopulation. Would like any step in the right direction.
To re-state, I have a Cox model with continuous predictors. I would like to build a Cox model but not stratify on rx but have this in the model. Then, I want to pass the created Cox object into ggadjustedcurves() function with uses "subpopulation re-balancing" when given a reference data set. And then, instead of showing two survival curves stratified on a categorical variable, I want to plot two representative survival curves at the 20th and 80th percentile.
EDIT
My first attempt
fit2 <- coxph( Surv(stop, event) ~ size + rx, data = bladder ) #remove strata
fit2
# CGP
pred<- data.frame("rx" = 1, "size" = 3.2)
ggadjustedcurves(fit2, data = pred , method = "conditional", reference = bladder)
Is this what I think it is? Conditional re-balancing has been applied to the reference data set and then the predicted curves are generated for an individual with rx=1 and size of 3.2.
It is difficult to understand what you are truly looking for, but I think I have a rough idea. I think you want to plot the survival curve that would have been observed if every person in your sample had received a specific value for the continuous covariate. If there is no confounding, you can simply use a Cox model that includes only the continuous covariate and use the predict() function for a range of points in time and plot the results. If you need to adjust for confounding, you can include the confounders in the Cox model and use g-computation to obtain the desired probabilities. I describe this in a recent preprint: https://arxiv.org/pdf/2208.04644.pdf
This can be done in R using the contsurvplot package (also developed by me). First, install the package using:
devtools::install_github("RobinDenz1/contsurvplot")
Afterwards, fit your Cox model, but use x=TRUE in the coxph call:
library(survival)
library(contsurvplot)
library(riskRegression)
library(ggplot2)
fit2 <- coxph(Surv(stop, event) ~ size + rx, data=bladder, x=TRUE)
You can now call the plot_surv_lines function to obtain the causal survival curves for specific values of size, given the model. Using the horizon argument you can tell the function for which values you want to plot the survival curves. I choose the 20% and 80% quantile of size as you described:
plot_surv_lines(time="stop",
status="event",
variable="size",
data=bladder,
model=fit2,
horizon=quantile(bladder$size, probs=c(0.2, 0.8)))
The package contains a lot more plotting routines to visualize the causal effect of a continuous variable on a time-to-event outcome that might be more suitable for what you actually want.
I am conducting a meta-analysis using the robust variance estimation (RVE) technique due to the fact that each study contains multiple effect sizes. In my case, effect sizes are r (linear association).
After extensive online research, I decided to use robumeta package in R, and robu function to calculate the overall effect size across all studies. Below is my main model and data structure in R.
`run.average <- robu(formula = Correlation ~ 1,
var.eff.size = Varience,
data = d2,
studynum = ID,
modelweights = "CORR")`
My goal is to create a forest plot to display the weighted mean effect size for each study (so, each study only has 1 effect size).
As far as I know, forest.robu() can plot each effect size, which, however, is not what I need. Using other functions like forest() may not apply to RVE model.
Therefore, I wonder if there is any solution to creating a forest plot for weighted mean effect size for each study.
I am conducting an analysis of where on the landscape a predator encounters potential prey. My response data is binary with an Encounter location = 1 and a Random location = 0 and my independent variables are continuous but have been rescaled.
I originally used a GLM structure
glm_global <- glm(Encounter ~ Dist_water_cs+coverMN_cs+I(coverMN_cs^2)+
Prey_bio_stand_cs+Prey_freq_stand_cs+Dist_centre_cs,
data=Data_scaled, family=binomial)
but realized that this failed to account for potential spatial-autocorrelation in the data (a spline correlogram showed high residual correlation up to ~1000m).
Correlog_glm_global <- spline.correlog (x = Data_scaled[, "Y"],
y = Data_scaled[, "X"],
z = residuals(glm_global,
type = "pearson"), xmax = 1000)
I attempted to account for this by implementing a GLMM (in lme4) with the predator group as the random effect.
glmm_global <- glmer(Encounter ~ Dist_water_cs+coverMN_cs+I(coverMN_cs^2)+
Prey_bio_stand_cs+Prey_freq_stand_cs+Dist_centre_cs+(1|Group),
data=Data_scaled, family=binomial)
When comparing AIC of the global GLMM (1144.7) to the global GLM (1149.2) I get a Delta AIC value >2 which suggests that the GLMM fits the data better. However I am still getting essentially the same correlation in the residuals, as shown on the spline correlogram for the GLMM model).
Correlog_glmm_global <- spline.correlog (x = Data_scaled[, "Y"],
y = Data_scaled[, "X"],
z = residuals(glmm_global,
type = "pearson"), xmax = 10000)
I also tried explicitly including the Lat*Long of all the locations as an independent variable but results are the same.
After reading up on options, I tried running Generalized Estimating Equations (GEEs) in “geepack” thinking this would allow me more flexibility with regards to explicitly defining the correlation structure (as in GLS models for normally distributed response data) instead of being limited to compound symmetry (which is what we get with GLMM). However I realized that my data still demanded the use of compound symmetry (or “exchangeable” in geepack) since I didn’t have temporal sequence in the data. When I ran the global model
gee_global <- geeglm(Encounter ~ Dist_water_cs+coverMN_cs+I(coverMN_cs^2)+
Prey_bio_stand_cs+Prey_freq_stand_cs+Dist_centre_cs,
id=Pride, corstr="exchangeable", data=Data_scaled, family=binomial)
(using scaled or unscaled data made no difference so this is with scaled data for consistency)
suddenly none of my covariates were significant. However, being a novice with GEE modelling I don’t know a) if this is a valid approach for this data or b) whether this has even accounted for the residual autocorrelation that has been evident throughout.
I would be most appreciative for some constructive feedback as to 1) which direction to go once I realized that the GLMM model (with predator group as a random effect) still showed spatially autocorrelated Pearson residuals (up to ~1000m), 2) if indeed GEE models make sense at this point and 3) if I have missed something in my GEE modelling. Many thanks.
Taking the spatial autocorrelation into account in your model can be done is many ways. I will restrain my response to R main packages that deal with random effects.
First, you could go with the package nlme, and specify a correlation structure in your residuals (many are available : corGaus, corLin, CorSpher ...). You should try many of them and keep the best model. In this case the spatial autocorrelation in considered as continous and could be approximated by a global function.
Second, you could go with the package mgcv, and add a bivariate spline (spatial coordinates) to your model. This way, you could capture a spatial pattern and even map it. In a strict sens, this method doesn't take into account the spatial autocorrelation, but it may solve the problem. If the space is discret in your case, you could go with a random markov field smooth. This website is very helpfull to find some examples : https://www.fromthebottomoftheheap.net
Third, you could go with the package brms. This allows you to specify very complex models with other correlation structure in your residuals (CAR and SAR). The package use a bayesian approach.
I hope this help. Good luck
I was wondering if I get some advice about fitting hurdle models using continuous data and covariates.
I have some continuous data that are generally well fit using a right-skewed distribution such as a Pareto, Gamma, or Weibull distribution. However, there several zeros in my data which are important to my analysis. In addition, I have some categorical (two-level) covariates and would like to model the parameters of a distribution as a function of these covariates in order to formally evaluate their importance (e.g., using AIC).
I have seen examples of hurdle models fit using continuous data but have not yet found any examples of how to incorporate covariates and a model-selection framework. Does anyone have any suggestions as to how to proceed or know of any R packages that allow this procedure? I have included some code below to reproduce the type of data I am working with. The non-zero data are generated via a generalized Pareto distribution from the package texmex. The parameters were estimated directly from my non-zero data. I have also included the code to plot the data in a histogram to see their distribution.
library("texmex")
set.seed(101)
zeros <- rep(0,8)
non_zeros <- rgpd(17, sigm=exp(-10.4856), xi=0.1030, u = 0)
all.data <- c(zeros,non_zeros)
hist(non_zeros,breaks=50,xlim=c(0,0.00015),ylim=c(0,9),main="",xlab="",
col="gray")
hist(zeros,add=TRUE,col="black",breaks=100,xlim=c(0,0.00015),ylim=c(0,9))
legend("topright",legend=c("zeros"),col="black",lwd=8)