Hacktoberfest 2026:維護者為十月標記出來的 issue,仍然開放、適合新手。 瀏覽 Hacktoberfest issue

Incorporate weightmask (per-exposure weight maps + masks) before the second v2 run

未關閉
#970 0 則留言 0 個 reaction 已指派 0 人 在 GitHub 檢視

維護者通常 1 天內回覆

還沒有人認領這個 Issue。

評估

難度
4/5
預估耗時
3-5 天
新手友好度
54/100
Issue 類型
功能
描述清晰度
基本清楚
活躍度
活躍
技術堆疊
python
領域
data

研究方向

Read ShapePipe's per-exposure masking setup and identify how it feeds the second v2 run; the issue does not name the relevant files or tests. Then review the weightmask package and the offered DRAC setup help to establish how to integrate it. Done means the second run uses weightmask masks and weight maps, enabling a tile or patch comparison with current masks.

由索引模型根據 Issue 內容生成。

描述

Seb Fabbro has written astroai/weightmask, a package that builds a weight map and a mask for FITS exposures. In his words, it is "fairly classic" (unlike maximask):

  • tuned astroscrappy for cosmics
  • use flatfield image for proper inverse variance map, and masking bad columns and hot pixels
  • masked saturated (not just based on a falsly reported saturation levels but on pixel histograms)
  • streaks masking: using hough/radon and other transforms. this part is quite slow though and not as robust, but you can disable it.

Seb says it would be easy to run on individual UNIONS exposures, and he has offered to help set it up for ShapePipe on the DRAC clusters.

Goal: add weightmask to ShapePipe's per-exposure masking before the second ShapePipe v2 run. It won't make the first run. The two runs then give a tile/patch-level comparison of the current masks against weightmask.

The histogram-based saturation mask is relevant to #943 (saturated pixels reaching ngmix unflagged).

🤖 Generated with Claude Code

主要語言
Python
星號
18
分支
14
平均合併
4 天 14 小時
30 天內合併 PR
44

環境準備

從這裡開始

  1. 先讀完整個 Issue,再讀專案的貢獻指南。
  2. 在 Issue 下留言說明你要接手 —— 這能避免兩個人做同樣的事。
  3. Fork 儲存庫,在一個分支上完成修改。
  4. 送出 Pull Request,並在描述裡引用這個 Issue 編號。

CosmoStat/shapepipe 的其他 Issue

查看 CosmoStat/shapepipe 的全部 Issue

相似的 Issue

更多 Python Issue

把新 issue 寄到你的電子郵件信箱

精選適合新手參與的 GitHub issue 摘要。