fix: Create indices with object dtype

[effigies]

Pandas is currently generating FutureWarnings:

.../.venv/lib/python3.12/site-packages/bids/variables/variables.py:535: FutureWarning: Setting an item of incompatible dtype is deprecated and will raise in a future error of pandas. Value 'False' has dtype incompatible with float64, please explicitly cast to a compatible dtype first.
  df.iloc[prev_ix:prev_ix + reps, col_ix] = v

pybids/bids/variables/variables.py

Lines 527 to 538 in 03a1af5

	def _create_index(all_keys, all_reps, all_ents):
	all_keys = np.array(sorted(all_keys))
	df = pd.DataFrame(np.zeros((sum(all_reps), len(all_keys))), columns=all_keys)
	prev_ix = 0
	for i, reps in enumerate(all_reps):
	for k, v in all_ents[i].items():
	col_ix = np.where(all_keys == k)[0][0]
	df.iloc[prev_ix:prev_ix + reps, col_ix] = v
	prev_ix = reps
	return df

Debugging, the issue is that we're creating the dataframe with a np.zeros(), so dtype=float, but entities and metadata may be strings, lists, etc. Using dtype=object resolves this.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.75%. Comparing base (03a1af5) to head (31f0441).
Report is 5 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1084   +/-   ##
=======================================
  Coverage   89.75%   89.75%           
=======================================
  Files          63       63           
  Lines        7123     7123           
  Branches     1363     1364    +1     
=======================================
  Hits         6393     6393           
  Misses        531      531           
  Partials      199      199           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[effigies]

Have resolved a few more warnings while I'm here.

@adelavega I'm done working on these for now. Care to review? Happy to explain changes more in case the commit messages and comments are insufficient.

[adelavega]

Why did we have this in the first place but is no longer necessary?

[effigies]

get_dummies() creates bools by default. This brought it back to float when reinserting into the dataframe. Instead, I make them floats at the beginning, since there's nothing that depends on them being bools above.

[adelavega]

Looks good, just one question