Support Series and DataFrame with DatetimeIndex

[Peque]

In example:

s = Series(
    index=date_range("2021", "2022", freq="MS", closed="left"),
    data=[-100] + [20] * 11,
)
xirr(s)  # 5.09623547168478

df = DataFrame(
    index=date_range("2021", "2022", freq="MS", closed="left"),
    data={
        "one": [-100] + [20] * 11,
        "two": [-80] + [19] * 11,
    },
)
xirr(df)  # Series(index=["one", "two"], data=[5.09623547168478, 8.780801977141174])

For the DataFrame I am suggesting returning a Series type simply because that is what i.e. df.sum() would return (applying .sum() to all columns).

Feel free to close if is out of scope for this module! (I would understand that) 😊

[Anexen]

Hi @Peque,
I can support the Series with DatetimeIndex and the case with DataFrame will be simply df.apply(xirr)

You can use the following code with the current implementation:

# for Series
xirr(s.reset_index())
xirr(s.index, s)

# for DataFrame
df.apply(lambda col: xirr(col.reset_index()))
df.apply(lambda col: xirr(col.index, col)

[Peque]

@Anexen Do you think it would make sense to return NaN instead of raising exceptions?

This would allow for df.apply(xirr) to work fine when possible.

In example:

from pandas import DataFrame
from pandas import date_range
from pyxirr import xirr

timezone = "Europe/Madrid"
index = date_range("2000-01-01", periods=10, freq="Y", closed="left", tz=timezone)
df = DataFrame(index=index, data={"a": 10, "b": 10})
df.iloc[0]["a"] = -25
df.apply(xirr)

Would return 0.3774630471754315, NaN.

[Anexen]

@Peque , I don't think so. I would like a clear separation between "Invalid input" case and "does not converge" case.

You can use something like this:

def suppress(func):
    def wrapper(*args, **kwargs):
        try:
            return func(*args, **kwargs)
        except InvalidPaymentsError:
            return None
    return wrapper
    
df.apply(suppress(xirr))