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.

library(h2o)
h2o.init()

# Upload the iris dataset
path <- "http://h2o-public-test-data.s3.amazonaws.com/smalldata/iris/iris_wheader.csv"
df <- h2o.importFile(path)

# Replace a single numerical datum. Note that columns and rows start at 1,
# so in the example below, the value in the 14th row and 2nd 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"] <- 4.4, "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"])

import h2o
h2o.init()
path = "http://h2o-public-test-data.s3.amazonaws.com/smalldata/iris/iris_wheader.csv"
df = h2o.import_file(path=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 first column by 3.
df[0] = 3*df[0]

# Replace by row mask. The example below searches for value less than 4.6 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"] = (df["sepal_len"] < 4.6).ifelse(4.6, df["sepal_len"])

# Replace missing values with 0.
df[df["sepal_len"].isna(), "sepal_len"] = 0

# Alternative with ifelse. Note the parantheses.
df["sepal_len"] = (df["sepal_len"].isna()).ifelse(0, df["sepal_len"])