R - 为什么 data.frame 将仅由数字组成的列转换为字符向量？

Question

Here is a simple dataframe I made, the data is originally in a pdf which is why I entered the column names first and then the rows like so.

skupni_indeksi <- data.frame(Proizvod =NA, Mjerna_jedinica=NA,Količine_2007=0,Količine_2008=0,Cijene_2007=0,Cijene_2008=0,stringsAsFactors = F)
skupni_indeksi[1,] <- c("A", "komad", 10, 12, 25, 30)
skupni_indeksi[2,] <-c("B", "litra", 30, 40, 40, 40)
skupni_indeksi[3,] <-c("C", "m2", 20, 15, 60, 72)

I know there are other ways of inputting such data, but my question concerns the behaviour of dataframes. Look at the column Cijene_2007, for example. It is originally a single number, and the new rows all place numbers into that column. Why is the column of a character type, when checking with str(skupni_indeksi)?

I found a manual solution to turn desired columns into numeric, simply:

skupni_indeksi[,3:6] <- sapply(skupni_indeksi[,3:6], as.numeric)

which fixes the issue. Could I have done something different in the first place to not need this line?

Answer 1

c("A", "komad", 10, 12, 25, 30) is a vector. A vector can have only one class.

Since you have numbers and characters mixed it turns the number to characters.

class(c("A", "komad", 10, 12, 25, 30))
#[1] "character"

You are filling the dataframe in a row-wise fashion. Usually, it is better to fill them in column-wise so that such data conversion do not take place.