feather
Fast on-disk format for data frames.
Read/Write
library(feather)
# Write feather file
write_feather(df, "data.feather")
# Read feather file
df <- read_feather("data.feather")
# Read specific columns
df <- read_feather("data.feather", columns = c("col1", "col2"))
Metadata
# Get metadata without reading data
meta <- feather_metadata("data.feather")
meta$dim # Dimensions
meta$types # Column types
meta$path # File path
With Arrow
library(arrow)
# Modern replacement using arrow
write_feather(df, "data.feather")
df <- read_feather("data.feather")
# Arrow table
tbl <- arrow_table(df)
write_feather(tbl, "data.feather")
Performance Tips
# Feather is column-oriented
# Best for: wide data, column selection
# Fast for: reading subsets of columns
# Read only needed columns
df <- read_feather("data.feather",
columns = c("id", "value"))
Cross-Language
# Python (pandas)
import pandas as pd
df = pd.read_feather("data.feather")
df.to_feather("data.feather")