Just some tips:

options(error=recover): it will tell R to launch debug section and you can choose which one to debug

options(show.error.locations=TRUE): let R show the source line number

Something else:

use traceback() to locate where the last error message is and then use browser() to run the function again to check what is wrong.

## Monday, December 30, 2013

## Tuesday, December 24, 2013

### A note about what is the output in predict.lm(lm_model, test_data, type=”terms”):

```
## first explain what is type="terms": If type="terms" is selected, a matrix of predictions
## on the additive scale is produced, each column giving the deviations from the overall mean
## (of the original data's response, on the additive scale), which is given by the attribute "constant".
set.seed(9999)
x1=rnorm(10)
x2=rnorm(10)
y=rnorm(10)
lmm=lm(y~x1+x2)
predict(lmm, data=cbind(x1,x2,y), type="terms")
lmm$coefficient[1]+lmm$coefficient[2]*x1+mean(lmm$coefficient[3]*x2)-mean(y)-predlm[,1]
lmm$coefficient[1]+lmm$coefficient[3]*x2+mean(lmm$coefficient[2]*x1)-mean(y)-predlm[,2]
```

## Monday, December 23, 2013

### R: Calculate ROC and Plot ROC

```
library(ROCR)
library(Hmisc)
## calculate AUC from the package ROCR and compare with it from Hmisc
# method 1: from ROCR
data(ROCR.simple)
pred=prediction(ROCR.simple$prediction, ROCR.simple$labels)
perf=performance(pred, 'tpr', 'fpr') #true positive and false negative
plot(perf, colorize=T)
perf2=performance(pred, 'auc')
auc=unlist(slot(perf2, 'y.values')) # this is the AUC
# method 2: from Hmisc
rcorrstat=rcorr.cens(ROCR.simple$prediction, ROCR.simple$labels)
rcorrstat[1] # 1st is AUC, 2nd is Accuracy Ratio(Gini Coefficient, or PowerStat, or Somer's D)
```

Subscribe to:
Posts (Atom)