Hi everyone,

I am trying to evaluate the explanatory character of a set of contextual variables over the efficiency levels using a bias-corrected data envelopment analysis. To do so, I am using the function dea.env.robust from rDEA package. One of the contextual variables (v9) is categorical and can assume 4 different values, which I have converted to 4 dummy variables ( 'v9_1','v9_2', 'v9_3' and 'v9_4').

I wrote:

```
library(rDEA)
data <- read.table("xxxxx.csv",sep=",", header = TRUE, dec='.')
inputs:: X = data[c('v1', 'v2','v3')]
outputs:: Y = data[c('v4')]
contextual variables:: Z = data [c('v5','v6,'v7','v8','v9_1','v9_2', 'v9_3','v9_4')]
```

Bias-corrected DEA score in an input-oriented model with constant returns-to-scale and environmental variables

```
data_env = dea.env.robust(
X= data[c('v1', 'v2','v3')] , Y= data[c('v4')] , Z= data [c('v5','v6,'v7','v8', 'v9_1','v9_2', 'v9_3', 'v9_4' )] , model="input",
RTS="constant",L1=100, L2=2000, alpha=0.05))
```

And then the following error message:

```
Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x), :
'data' must be of a vector type, was 'NULL'
```

I have been looking for help on several websites and blogs but I did not find a solution. I know that if I do not consider the last dummy variable (v9_4), it works but I do not understand why.

The output of the dea.env.robust function does not return p-values nor z-scores values of the estimated coefficients. Then I do not know how to evaluate if beta coefficients obtained are significant or not.

Could someone please help me with these two questions?

Thanks in advance

Ana Loureiro