Hello! I am retrieving the embeddings generated for categorical variables for using as an input in another model. I know the order of embeddings is the same as that of the categorical variables that we specify, but I’m a little confused of the order of the values within each embedding.
For example, let’s say I have a DataFrame with a categorical variable (‘thing’) like the following:
imp thing
0 1 fish
1 2 cake
2 3 melon
3 7 melon
After converting it to categories, it gives me:
Categories (3, object): [cake < fish < melon]
Now, suppose I have the following embedding for the variable:
[0.2. 0.4
0.1 0.3
0.9. 0.7]
Is the order same as the order of the values, i.e. does this mean that [0.2 0.4] is the embedding for cake, [0.1 0.3] for fish, and [0.9 0.7] for melon? Or is the order according to the order of the values in the data frame, i.e. [0.2 0.4] for fish, [0.1 0.3] for cake and so on?
Thanks in advance!