Special handling of inputs

Unlike R's magrittr, dato allows for any number of arguments to be passed through >>, and used in order as arguments in the receiving Pipeable object. While this can be implemented using some simple array-processing logic, we provide a number of decorators and arguments to make this process more convenient:

  • unpack_input: pass input as *input.

  • : pass only the first element of input.

For example, a common example is the pandas.merge method, which merges two tables

df = pd.merge(a, b, on='id') # pandas implementation.
(a, b) >> Merge # dato implementation.

The above can be implemented without the unpack_input decorator as follows:

def Merge(df_tuple, args, **kwargs):
a = df_tuple[0] b = df_tuple[1]
return pd.merge(a, b, args, **kwargs)

Or, more simply with unpack_input as:

@unpack_input
@Pipeable
def Merge(args, **kwargs):
pd.merge(args, **kwargs)

‚Äč