special-data-types
Use Lux semantic data types, temporal and geographic handling, ID suppression, and index/grouped dataframe recommendations.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- special-data-types
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Use Lux semantic data types, temporal and geographic handling, ID suppression, and index/grouped dataframe recommendations.
- disable-model-invocation
- true
02 资源文件 3
- 参考文档references/data-types.md7.9 KB
- 参考文档references/troubleshooting.md7.2 KB
- 脚本scripts/data_type_smoke.py4.5 KB
03 SKILL.md 正文
Lux Special Data Types
Use this sub-skill when a Lux workflow depends on the semantic meaning of columns rather than only Pandas dtypes: inferred or overridden data types, temporal columns, geographic columns, identifier suppression, named indexes, and grouped or crosstabbed dataframes.
Route here for
- Inspecting
df.data_typeand correcting Lux semantic types withdf.set_data_type(...). - Explaining
quantitative,nominal,geographical,temporal, andidcolumns. - Fixing date-like columns so Lux uses temporal line charts and time-aware recommendations.
- Preparing
stateorcountrycolumns for geographic recommendations. - Understanding why ID-like fields are not visualized.
- Working with named row or column indexes, groupby aggregations, pivots, and crosstabs that produce
Row GroupsorColumn Groupsrecommendations. - Diagnosing empty, very small, or hierarchical-index dataframe messages.
Route elsewhere
- For general Pandas import order,
df.intent, recommendation keys, and default/intent action behavior, usepandas-intent-recommendations. - For
Clause,Vis,VisList, chart marks, Altair/Matplotlib/Vega-Lite export, or exported code, usevisualization-export. - For global Lux config, widget setup, plotting backend/style, or custom recommendation actions, use
configuration-actions. - For PostgreSQL-backed
LuxSQLTableworkflows, usesql-backend.
Operating references
- Read
references/data-types.mdfor the self-contained data-type, temporal, geographic, ID, and index/group workflow guide. - Read
references/troubleshooting.mdwhen Lux gives a warning/message or when expected recommendations are missing. - Run
scripts/data_type_smoke.py --helporscripts/data_type_smoke.pyin an environment withlux-apiinstalled to verify the core offline data-type behaviors on tiny fixtures.
Minimal workflow reminder
- Import Lux before creating Pandas dataframes:
import lux; import pandas as pd. - Inspect
df.data_typeafter constructing or mutating a dataframe. - Convert real temporal columns with
pd.to_datetime(...); if a pre-existingViswas built from the old source, refresh it withvis.refresh_source(df). - Override semantic mistakes with
df.set_data_type({"column": "nominal"}), using only Lux's accepted override values. - Flatten unsupported hierarchical indexes with
df.reset_index()before asking Lux for recommendations.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).