Replacing Values in a Frame¶
This example shows how to replace numeric values in a frame of data. Note that it is currently not possible to replace categorical value in a column.
- r
- python
> library(h2o)
> path <- "data/iris/iris_wheader.csv"
> h2o.init(nthreads=-1)
> df <- h2o.importFile(path)
# Replace a single numerical datum. Note that columns and rows start at 0, so in the example below, the value in the 15th row and 3rd column will be set to 2.0.
> df[14,2] <- 2.0
# Replace a whole column. The example below multiplies all values in the second column by 3.
> df[,1] <- 3*df[,1]
# Replace by row mask. The example below searches for value less than 4.4 in the sepal_len column and replaces those values with 4.6.
> df[df[,"sepal_len"] < 4.6, "sepal_len"] <- 4.6
# Replace using ifelse. Similar to the previous example, this replaces values less than 4.6 with 4.6.
> df[,"sepal_len"] <- h2o.ifelse(df[,"sepal_len"] < 4.4, 4.6, df[,"sepal_len"])
# replace missing values with 0
> df[is.na(df[,"sepal_len"]), "sepal_len"] <- 0
# alternative with ifelse
> df[,"sepal_len"] <- h2o.ifelse(is.na(df[,"sepal_len"]), 0, df[,"sepal_len"])