File size: 2,880 Bytes
5995ef5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
[project]
name = "macrolens"
version = "0.1.0"
description = "MacroLens: a contextual financial benchmark (NeurIPS 2026 D&B submission)"
readme = "README.md"
license = { text = "MIT" }
requires-python = ">=3.11, <3.12"

dependencies = [
  # Core scientific
  "numpy>=2.2.6",
  "pandas>=2.3.0",
  "pyarrow>=22.0.0",
  "scipy>=1.13",
  "scikit-learn>=1.5",
  "tabulate>=0.9.0",
  "tqdm>=4.67.1",
  "joblib>=1.4",
  "pyyaml>=6.0.2",
  # HTTP / I/O
  "httpx>=0.28.1",
  "python-dotenv>=1.1.1",
  "beautifulsoup4>=4.13.4",
  "html2text>=2025.4.15",
  # Tabular / TS classical
  "lightgbm>=4.5",
  # Deep learning core
  "torch>=2.5",
  "transformers>=4.45",
  "accelerate>=1.0",
  "huggingface-hub>=0.30",
  "safetensors>=0.5",
  "datasets>=3.0",
  "einops>=0.8",
  "sentence-transformers>=4.0",
  "timm>=1.0",
  # TSFM ZS panel
  "chronos-forecasting>=2.0",
  "timesfm>=1.2",
  # Data sources
  "yfinance>=0.2.65",
  # Lakehouse stores
  "duckdb>=1.0",
  "neo4j>=5.20",
  "pymilvus>=2.4",
  "milvus-lite>=2.4",
  # Agents / LangChain
  "langchain>=0.3",
  "langchain-core>=0.3",
  "langchain-community>=0.3",
  "langchain-openai>=0.3",
  "langgraph>=0.2",
  "openai>=1.50",
  # Plot / explorer
  "matplotlib>=3.9",
  "plotly>=5.24",
  "jinja2>=3.1",
  # Tests
  "pytest>=8.3",
  "pytest-asyncio>=1.0",
  "pytest-cov>=6.1",
]

[project.optional-dependencies]
vllm = ["vllm>=0.7"]
qlora = ["peft>=0.13", "trl>=0.12", "bitsandbytes>=0.44"]
# Moirai 2.0 (uni2ts) needs lightning + gluonts at runtime, but uni2ts 2.0.0
# itself ships with stale scipy<1.12 / torch<2.5 caps. Install order:
#   uv pip install -e ".[vllm,qlora,explorer,moirai-runtime]"
#   uv pip install --no-deps uni2ts==2.0.0
# `moirai-runtime` provides the runtime libs uni2ts uses; the --no-deps
# install adds uni2ts itself without re-resolving its stale upper bounds.
moirai-runtime = [
  "lightning>=2.4",
  "gluonts>=0.16",
  "jaxtyping>=0.2",
]
explorer = [
  "dash>=3.2",
  "dash-bootstrap-components>=2.0",
  "dash-mantine-components>=0.14",
  "dash-ag-grid>=31.0",
  "dash-iconify>=0.1",
  "streamlit>=1.45",
]

[build-system]
requires = ["hatchling"]
build-backend = "hatchling.build"

[tool.hatch.build.targets.wheel]
# Explicit packages -- avoids setuptools auto-discovery confusion across
# the many top-level dirs (data/, data_lakehouse/, data_small_caps/, etc.).
packages = [
  "macrolens",
  "dataloader",
  "agents",
  "baselines",
  "experiments",
  "lakehouse",
  "explorer",
]

[tool.uv]
prerelease = "allow"

# Override deps where upstream pins are stale.
# - uni2ts 2.0 declares `torch>=2.1,<2.5`, but it works on torch 2.5+ in
#   practice (no breaking torch-API usage); we need torch>=2.5 for vllm.
override-dependencies = [
  "torch>=2.5",
]

[tool.pytest.ini_options]
asyncio_mode = "auto"
testpaths = ["tests"]
python_files = ["test_*.py"]
filterwarnings = [
  "ignore::DeprecationWarning",
]