Refactor timeseriesservice

[BytecodeBrewer]

What changed?

• seperated functionalies in time_series_service into two new services: market_data_service and trend_analysis_service
• restructured the workflow between service layer and storage layer as well as service layer and client layer
• edit the domain models and add a scheme for price bar instead of a model
• implemented the first storage logic
• trend_analysis_service calls analytics not vise versa
• add and edit a lot of tests and improved the test qulity

Related issue

Closes #70

Tests

• I ran pytest
• Existing tests pass
• I added or updated tests where necessary
• Not needed because this only affects documentation or repository setup

Checklist

• The change is focused
• No secrets or API keys are included
• Documentation was updated if needed
• CI is expected to pass

Notes

• I know that I made more than the tickets said but some things before were really messed up and weren't connected good enough. Though I had to fix the cascade of tests

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 469c45833d

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Import the trend service from the module that defines it

Launching the GUI now fails during import because market_data_service only defines get_market_data; the new prepare_trend_analysis function is defined in argus.services.trend_analysis_service. In any environment with the declared dependencies installed, importing argus.gui.app or argus.main raises ImportError before the app window can start.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Convert price bars before running trend metrics

When get_market_data succeeds, response.bars uses the new price-bar schema (timestamp/close), but the existing trend metric functions still index rate and later date. A normal successful trend request therefore raises KeyError: 'rate' here instead of returning a figure; this was masked because the added trend-analysis tests are inside a module-level string.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Keep timeframe in stored price-bar identity

The request model still carries timeframe, and get_market_data checks storage before calling yfinance, but the table no longer stores or keys by timeframe. After caching daily bars for an instrument/date range, a later hourly request for the same source/instrument/dates will be treated as a storage hit and return the daily rows instead of fetching hourly data.

Useful? React with 👍 / 👎.