feat: 3-part make for LFP

[ttngu207]

This pull request refactors the make method in element_array_ephys/ephys_no_curation.py to improve modularity and readability by splitting it into three distinct methods: make_fetch, make_compute, and make_insert. Additionally, it introduces class-level constants for better code organization and replaces hardcoded values with these constants.

Refactoring for Modularity:

• Split the original make method into three separate methods:
  • make_fetch: Handles fetching data and returns relevant information, such as file paths, LFP indices, probe information, and electrode data. [1] [2]
  • make_compute: Processes the fetched data to compute broadband LFP signals for each electrode, including filtering and downsampling. Returns computed LFP signals and mapping data.
  • make_insert: Inserts computed data and metadata into the database, ensuring separation of concerns.

[Copilot]

Pull Request Overview

This PR refactors the LFP processing pipeline by splitting the original "make" method into three distinct methods: "make_fetch", "make_compute", and "make_insert".

• Split the original make function into three parts to separate fetching, computing, and inserting operations.
• Replaced usages of local constants with class attributes (e.g., TARGET_SAMPLING_RATE, POWERLINE_NOISE_FREQ, MAX_DURATION_MINUTES).

Comments suppressed due to low confidence (2)

element_array_ephys/ephys_no_curation.py:412

• The variable 'channels' is used in the 'make_insert' method but is not defined or passed as a parameter. Consider passing 'channels' to 'make_insert' or computing it within this method.

for ch_idx, lfp in zip(channels, all_lfps):

element_array_ephys/ephys_no_curation.py:264

• The method now relies on 'self.MAX_DURATION_MINUTES', 'self.TARGET_SAMPLING_RATE', and 'self.POWERLINE_NOISE_FREQ' instead of local constants; ensure these class attributes are defined to prevent runtime errors.

duration <= self.MAX_DURATION_MINUTES