BUG: Fix Backtest() with a non-datetime index on pandas 3

[David-Wu1119]

On pandas 3.0, Backtest(data, ...) raises for any index that is neither a DatetimeIndex nor a RangeIndex:

import pandas as pd
from backtesting import Backtest, Strategy
from backtesting.test import GOOG

class Noop(Strategy):
    def init(self): pass
    def next(self): pass

data = GOOG.copy()
data.index = data.index.strftime('%Y-%m-%d')   # or epoch integers
Backtest(data, Noop)
# AttributeError: 'Index' object has no attribute 'is_numeric'

Index.is_numeric() was deprecated in pandas 2.0 and removed in 3.0. setup.py allows pandas >= 0.25, so a fresh install gets pandas 3. The conversion branch below it also passed infer_datetime_format=True, which 3.0 removed too. That raises TypeError, which the surrounding except ValueError doesn't catch.

This PR:

• replaces data.index.is_numeric() with pd.api.types.is_numeric_dtype(data.index), which exists in every supported pandas version;
• drops infer_datetime_format=True. It only affected string parsing, and this branch only handles numeric input.

The existing suite uses GOOG's DatetimeIndex, which short-circuits before the removed call. That's why CI on pandas 3 didn't catch this.

Test: added TestRegressions.test_numeric_and_string_index:

• An epoch-nanosecond integer index is still converted to the original DatetimeIndex.
• A date-string index still gets the "index is not datetime" warning instead of an exception.

On master with pandas 3.0.6 the test fails with the AttributeError above.

Verification:

• python -m backtesting.test: 82 tests pass with 1 skipped, on both pandas 3.0.6 and pandas 2.2.3.
• flake8 backtesting is clean.
• mypy backtesting reports the same 8 errors as master.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.