Gaps found by the Panel outputs prototype
维护者通常 1 天内回复
还没有人认领这个 Issue。
评估
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 新手友好度
- 45/100
- Issue 类型
- 功能
- 描述清晰度
- 基本清楚
- 活跃度
- 活跃
- 技术栈
- jupyter, python
调研方向
Start with docs/developer/roles/spec-form.ipynb on the spec-form-outputs branch and inspect the five gaps recorded in the issue. The issue says the prototype works around them and identifies #81 for the completion notification; first determine which gap is in scope. Done means addressing a selected gap, with the prototype behavior improved as described.
由索引模型根据 Issue 内容生成。
描述
Gaps found while adding an outputs list and plotting to the Panel prototype (docs/developer/roles/spec-form.ipynb, branch spec-form-outputs). The app lists a row per output of each record, shows what the output is and how it was made, and plots arrays with plopp. None of these block the prototype.
1. A value's summary reads the whole value (moved to #149)
To show an output's sizes, unit and coordinates, the app calls client.output(record, name), which copies the whole value. No call returns this without the data:
value = client.output(record, 'counts') # copies the whole array
f'{value.dims} {value.shape} [{value.unit}]'
In this process the cost is small. For the hosted service (#27), listing or inspecting outputs would transfer every value.
2. The spec doesn't declare an array's number of dimensions (moved to #149)
The spec declares sc.Variable or sc.DataArray, but not how many dimensions. ForwardOutputs.i0 is declared as an array, so the app offers Plot, but its value is 0-D. The app only finds out after reading it and shows "0-D, not plotted". The same holds for arrays with more than 2 dimensions.
3. Nothing tells the app that a record finished (moved to #148)
The app submits without waiting, then polls client.status every 0.5 s through a Panel periodic callback (App.poll). The missing as_completed / timeout= is already listed in #81. In the prototype, each finished record also rebuilds the pickers of every form, since a form offers each record as a reference.
4. A workflow can't learn a run's number from its file model
A file's identity is opaque to the workflow. toy_sans.COUNTS outputs a (run, pixel) array, but cannot give the run dimension a coordinate with the run numbers, so the plot labels runs by index.
5. Default titles from field names read poorly
places.title turns iofq into "Iofq". The spec author can set Field(title=...), so this needs nothing from the framework, but the toy specs and examples should set titles (and description=) on outputs.
- 主要语言
- Python
- 星标
- 0
- 派生
- 1
- 平均合并
- 9 小时 40 分钟
- 30 天内合并 PR
- 60
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
scipp/essapps 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 62/100
维护者通常 1 天内回复
-
A number that is not finite is stored as None at a field that accepts None可能已有人在做 @jeevanabhi-r 于 3 天前认领。 未关闭bug
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
-
难度 5/5 一周以上 新手友好度 25/100
维护者通常 1 天内回复
-
难度 5/5 一周以上 新手友好度 25/100
维护者通常 1 天内回复
-
难度 5/5 一周以上 新手友好度 25/100
维护者通常 1 天内回复
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 84/100
NVIDIA/auto-ontology#463 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 86/100
NousResearch/hermes-agent#151165 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 66/100
维护者通常 1 天内回复
-
code-health
难度 2/5 1-3 小时 新手友好度 78/100
tinaudio/synth-setter#3750 ·
维护者通常 1 天内回复
-
[Bug] vllm-sr config migrate drops providers.models[].deployment可能已有人在做 @Bevisy 今天认领。 未关闭needs-acceptance wg/developer-experience-ecosystem
难度 2/5 1-3 小时 新手友好度 86/100
vllm-project/semantic-router#5996 ·
维护者通常 2 天内回复