Predict function convert Probabilities between 0 to 1 and store in a dataframe

Question

The Predict function code returns output like this.

library(e1071)
model <- svm(Species ~ ., data = iris, probability=TRUE)
pred <- predict(model, iris, probability=TRUE)

head(attr(pred, "probabilities"))
#      setosa versicolor   virginica
# 1 0.9803339 0.01129740 0.008368729
# 2 0.9729193 0.01807053 0.009010195
# 3 0.9790435 0.01192820 0.009028276
# 4 0.9750030 0.01531171 0.009685342
# 5 0.9795183 0.01164689 0.008834838
# 6 0.9740730 0.01679643 0.009130620

So, I wrote a piece of code like this :-

Code:-

pred_df <- as.data.frame(pred)

This returns output like this, (have just made up the values)

1 Setosa
2 Versicolor
3 Virginica
4 Setosa
5 Setosa

But my preferred output would be something like this (have just made up the values),

   Setosa      Versicolor       Virginica
1 0.62          0.11               0.27
2 0.41          0.55               0.04

***Pred is a factor and Pred_df is a dataframe***

I am looking to return the numbers in the form of a decimal rather than a whole number. Kindly help me with this.

Are you aware that there is a difference between the internal representation of a number and how it is printed? You seem to care about the printing and should look into using functions like `format`, `formatC`, `prettyNum`, or `sprintf`. — Roland, Nov 28 '18 at 11:57
@Roland Pred is a factor and Pred_df is a dataframe & this conversion changes the format. No, I am not worried about printing. I want the figures to be in the form of a decimal number ranging from 0 to 1 to export. — Student of the Digital World, Nov 28 '18 at 11:58
The `augment` function in the `broom` package is a great option for doing things like this (i.e. getting prediction or model statistics into a tidy data.frame) — jmuhlenkamp, Nov 28 '18 at 12:02
@Roland No sir. Returns the output in the form of 0's and 1's. — Student of the Digital World, Nov 28 '18 at 12:10
Sir, please provide a reproducible example (see [this FAQ](https://stackoverflow.com/a/5963610/1412059)) if you require further assistance. Thank, you, sir. — Roland, Nov 28 '18 at 12:14
@Roland Will do. Just noticed I had my output wrongly mentioned in the question. fixed it now. — Student of the Digital World, Nov 28 '18 at 12:35

score 2 · Accepted Answer · answered Nov 28 '18 at 12:44

To see what's going on you need to look inside the structure.

str(pred)
 Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
 - attr(*, "names")= chr [1:150] "1" "2" "3" "4" ...
 - attr(*, "probabilities")= num [1:150, 1:3] 0.979 0.971 0.978 0.973 0.978 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : chr [1:150] "1" "2" "3" "4" ...
  .. ..$ : chr [1:3] "setosa" "versicolor" "virginica"

So

as.data.frame(attr(pred,"probabilities"))

should do what you want.

Predict function convert Probabilities between 0 to 1 and store in a dataframe

1 Answers1