mlforecast with DASK has incorrect/incomplete dependencies

[arizzuto]

What happened + What you expected to happen

installing an environment with mlforecast = {extras=["dask"], version="*"}

does not install all required dependencies. An extra line is required: dask = {extras=["distributed", "dataframe"], version="version matching whatever dask version is install above"}

Versions / Dependencies

[tool.poetry] name = "sort_arbitrage_settings_lambda" version = "0.0.1" description = "Sort arbitrage settings for a fan out taking just things set for the arbitrage optimiser" authors = ["Aaron Rizzuto"] package-mode = false

[tool.poetry.dependencies] python = "3.12." mlforecast = {extras=["dask"], version=""} dask = {extras=["distributed", "dataframe"], version="2024.12.1"}

Reproduction script

Running the dask example workflow on the nixtla website produces this result

Issue Severity

Medium: It is a significant difficulty but I can work around it.