Don't concatenate data for region_df / field_df before necessary

[hoxbro]

Save intermediate data for field_df and region_df in a list. This list is then concatenated into the respective dataframe when called. An example of code before and after can be run with this code:

This branch: Define annotations (size=5,000): 0.067 seconds
Main branch: Define annotations (size=5,000): 6.149 seconds

from contextlib import contextmanager
from time import perf_counter

import numpy as np
import pandas as pd

from holonote.annotate import Annotator, SQLiteDB


@contextmanager
def catchtime(message):
    start = perf_counter()
    yield
    print(f"{message}: {perf_counter() - start:.3f} seconds")


annotator = Annotator({"TIME": int}, connector=SQLiteDB(table_name="slowness"))


N = 5_000
rng = np.random.default_rng(1337)
start_time = np.arange(0, N * 2, 2)
end_time = start_time + 1
data = pd.DataFrame(
    {
        "start_time": start_time,
        "end_time": end_time,
        "description": rng.choice(["A", "B"], N),
    }
)

with catchtime(f"Define annotations (size={N:,})"):
    annotator.define_annotations(data, TIME=("start_time", "end_time"))

[hoxbro]

I don't think this is needed anymore but I'm not completely sure.

[hoxbro]

This class is not used as a parameterized class and makes it possible to remove the method _update_index for a property that looks up the index.

[codspeed-hq]

CodSpeed Performance Report

Merging #72 will improve performances by ×11

_{Comparing dont_concat_before_nessesary (94410c6) with main (a8562ed)}

Summary

⚡ 3 improvements

Benchmarks breakdown

	Benchmark	main	dont_concat_before_nessesary	Change
⚡	test_define_annotations[1000]	14,238.4 ms	405.1 ms	×35
⚡	test_define_annotations[100]	873.8 ms	43.1 ms	×20
⚡	test_define_annotations[10]	86.4 ms	7.6 ms	×11