Based on the Modified PME method.

from pypme import verbose_xpme
from datetime import date
pmeirr, assetirr, df = verbose_xpme(
    dates=[date(2015, 1, 1), date(2015, 6, 12), date(2016, 2, 15)],
    cashflows=[-10000, 7500],
    prices=[100, 120, 100],
    pme_prices=[100, 150, 100],
)

Will return 0.5525698793027238 and 0.19495150355969598 for the IRRs and produce this dataframe:

Notes:

• The cashflows are interpreted from a transaction account that is used to buy from an asset at price prices.
• The corresponding prices for the PME are pme_prices.
• The cashflows is extended with one element representing the remaining value, that's why all the other lists (dates, prices, pme_prices) need to be exactly 1 element longer than cashflows.

• xpme: Calculate PME for unevenly spaced / scheduled cashflows and return the PME IRR only. In this case, the IRR is always annual.
• verbose_xpme: Calculate PME for unevenly spaced / scheduled cashflows and return vebose information.
• pme: Calculate PME for evenly spaced cashflows and return the PME IRR only. In this case, the IRR is for the underlying period.
• verbose_pme: Calculate PME for evenly spaced cashflows and return vebose information.
• tessa_xpme and tessa_verbose_xpme: Use live price information via the tessa library. See below.

Use tessa_xpme and tessa_verbose_xpme to get live prices via the tessa library and use those prices as the PME. Like so:

from datetime import datetime, timezone
from pypme import tessa_xpme
common_args = {
    "dates": [
        datetime(2012, 1, 1, tzinfo=timezone.utc), 
        datetime(2013, 1, 1, tzinfo=timezone.utc)
    ],
    "cashflows": [-100],
    "prices": [1, 1],
}
print(tessa_xpme(pme_ticker="LIT", **common_args))  # source will default to "yahoo"
print(tessa_xpme(pme_ticker="bitcoin", pme_source="coingecko", **common_args))
print(tessa_xpme(pme_ticker="SREN.SW", pme_source="yahoo", **common_args))

Note that the dates need to be timezone-aware for these functions.

Note that the package will only perform essential sanity checks and otherwise just works with what it gets, also with nonsensical data. E.g.:

from pypme import verbose_pme
pmeirr, assetirr, df = verbose_pme(
    cashflows=[-10, 500], prices=[1, 1, 1], pme_prices=[1, 1, 1]
)

Results in this df and IRRs of 0:

• tessa: Find financial assets and get their price history without worrying about different APIs or rate limiting.
• strela: A python package for financial alerts.

• Google Sheet w/ reference calculation
• Modified PME on Wikipedia