Releases: GongJr0/NeoPortfolio
Backtest modules release
The release includes BtEngine
and BtStrategy
to enable backtesting capability with a pipeline being developed to automatically backtest optimized portfolios from Markowitz
and nCrOptimize
.
Error Handling and Cache Refactor
Increased the assertions and error check in all modules. Moved the caching modules to the single Cache.py
file and refactored the classes to have a layer of abstraction by inheriting common functions from the new CacheConstructor
class.
`nCrResult` Module release
This release includes the nCrResult
custom type as the return of the nCrOptimize.optimize_space
method. nCrResult
extends the standard python list
type and provides added functionality around visualization and report compilation regarding the outcome of the weight optimizations.
Automated Portfolio Selection
This release incorporates the nCrEngine
, nCrOptimize
, PortfolioCache
, and nCrCache
modules to allow access to the beta release of the combitation optimization functionality.
Sentiment bug fixes
Fixed an issue with FinBERT
label ordering. Switched labels={0: "negative", 1: "neutral", 2: "positive"}
with labels={0: "neutral", 1: "positive", 2: "negative"}
to follow convention.
First Release
This is the initial release of the NeoPortfolio
package that aims to extend the functionality of MPT and adjust the theoretical methodology from 1952 to todays standards through ML and automation.