What is the difference between read file through csv and pipe reading in julia?

user_9381966 0
Viewed 59

What is the difference between:

using CSV;

df = CSV.read("data.csv");

and

df = CSV.read("data.csv") |> DataFrame;

I mean, in the first case df is read as DataFrame, what the need for the pipe in the second case?

julia dataframe
2 Answers
df = CSV.read("data.csv");

is currently a shorthand for:

df = CSV.read("data.csv", DataFrame);

(in the future the former will be disallowed most likely)

I am not sure where you have seen:

df = CSV.read("data.csv") |> DataFrame;

as this is not needed. You create an unnecessary additional copy of your data frame.

Probably you meant:

df = CSV.File("data.csv") |> DataFrame;

which is a low-level API of CSV.jl.

In short. It is recommended to use:

df = CSV.read("data.csv", DataFrame);
user_12695670

It's just for explicitness. CSV.read("data.csv") |> DataFrame is identical to DataFrame(CSV.read("data.csv")), which in principle guarantees that you'll end up with a DataFrame even if CSV changes their output format at some point in the future. Note that it would be more ideomatic to write CSV.File("data.csv") |> DataFrame, and this makes it easy to change the output format at any time in the future.

user_25400370
Related