You can use feather.
It's a data format for data frames (created by @Wes McKinney and @hadley) to make data sharing between R and python easy (and some other languages too).
In R:
library(feather)
file_path <- "foo.feather"
data_frame <- read_feather(file_path)
write_feather(data_frame, file_path)
In python:
import feather
file_path = 'foo.feather'
data_frame = feather.read_dataframe(file_path)
feather.write_dataframe(data_frame, file_path)
PS.: Podcast on feather where authors discuss it's application, pros/cons and future.
