HomeArticlesModelViewerContact

Rules for healthy and happy Data

16 August, 2020 - 1 min read - Tags: Grasshopper,Youtube

やったこと

  • Andrew Heumann のデータツリーの動画見た

  • TAEC のスライドだいたいまとめた

    • Craneのパートはもうちょっとつめたい

技術メモ

  • Karambaはデータツリーで入力できることがわかったのでそんなに苦労しないで、HoaryFoxからKarambaへの出力できそう

    • 問題はデータフロー的に処理をインスタンス渡しにしたいので、ベースは書き直さないといけない

TIL

Andrew's Rules for healthy and happy Data Trees

  1. All paired branches should have either N items or 1 item. Don't mix lists of different lengths.

    • [0, 1, 2] + [0] and [0, 1, 2] + [a, b, c] are fine.
    • [0, 1, 2] + [0, 1] is not fine.
  2. All inputs should have either M branches or 1 branches. Don't mix inputs with different branch counts.

    • [{0;0}, {0;1}, {0;2}] + [0;0] is fine.
    • [{0;0}, {0;1}, {0;2}] + [{0;0}, {0;1}] is not fine.
  3. Look before you merge! No Jagged Data Trees ever!

    • [{0;0}, {0;1}] is fine.
    • [{0}, {0;0}, {0;1}] is not fine.
  4. Avoid destructive operations like flatten, simplify, and Path Mapper.They break your definition when things change upstream!

  5. Avoid Simplify. Instead, PShift early, PShift often.
  6. Design your scripts as though the upstream trees are going to change shape (because they will!)

その他

  • アンドリューの健康的でパッピーなデータツリーのルール という言い回し好き
  • HoaryFoxのKarambaコンバーターの実装、技術的には可能だがいろいろ負債の修正がめんどくさい

© 2024 hrntsm / Built with

Gatsby