vault backup: 2026-03-31 16:53:06

timestamped/2025-12-17_05-39-00.md

@@ -9,6 +9,8 @@ dg-publish: true
 ---
 # 2025-12-17 05:39:??
 
+%% [[statistical-modeling-for-construction-estimating]] %%
+
 One aspect of estimating that I find most interesting,
 but that is criminally understudied,
 is the effect of building dimensions

timestamped/2026-03-31_07-43-16.md

@@ -21,8 +21,76 @@ yearly: "[[2026]]"
 
 ### Cross Join
 
-[](https://www.sqlshack.com/sql-cross-join-with-examples/)
+In [[sql]], the `CROSS JOIN` operation
+returns the [cartesian product](https://en.wikipedia.org/wiki/Cartesian_product)
+of two tables.
 
-### ~~add column, update by key~~
+```sqlite
+CREATE TABLE WireSizes (wire_size TEXT NOT NULL);
+INSERT INTO WireSizes VALUES ('#12'), ('#10'), ('#8'), ('#6'), ('#4'), ('#2'), ('#1'), ('#1/0'), ('#2/0');
+
+CREATE TABLE WireMaterials (wire_material TEXT NOT NULL);
+INSERT INTO WireMaterials VALUES ('CU'), ('AL');
+
+SELECT wire_size, wire_material
+FROM WireSizes
+CROSS JOIN WireMaterials;
+```
+
+```sqlite
+SELECT wire_size||' '||wire_material AS wire_type
+FROM WireSizes, WireMaterials; -- implicit cross join
+```
+
+### Create Table From Query
+
+`CREATE TABLE _ AS` syntax is intuitive,
+
+```sqlite
+CREATE TABLE Wires AS
+SELECT wire_size, wire_material
+FROM WireSizes, WireMaterials;
+```
+
+but it excludes the ability to set constraints.
+
+Use this pattern instead:
+
+```sqlite
+CREATE TABLE Wires (
+    wire_size TEXT NOT NULL,
+    wire_material TEXT NOT NULL,
+    PRIMARY KEY (wire_size, wire_material),
+    FOREIGN KEY (wire_size) REFERENCES WireSizes(wire_size),
+    FOREIGN KEY (wire_material) REFERENCES WireMaterials(wire_material)
+);
+
+INSERT INTO Wires (wire_size, wire_material)
+SELECT wire_size, wire_material
+FROM WireSizes, WireMaterials;
+```
+
+You'd think you could just throw an `AS` after the constraints
+and omit the `INSERT` statement, but no.
+
+### Add Columns to Table
 
 for adding data after creating new tables with [[#Cross Join]]
+
+```sqlite
+ALTER TABLE Wires
+ADD length_specific_dc_resistance_in_milliohms_per_foot FLOAT;
+
+UPDATE Wires
+SET length_specific_dc_resistance_in_milliohms_per_foot = CASE
+   WHEN wire_size = '#8' AND wire_material = 'CU' THEN 0.778
+   WHEN wire_size = '#6' AND wire_material = 'CU' THEN 0.491
+   WHEN wire_size = '#4' AND wire_material = 'CU' THEN 0.308
+   -- ...
+   ELSE NULL
+END;
+```
+
+### Resources
+
+* [sqlshack.com](https://www.sqlshack.com/)

timestamped/2026-03-31_14-53-42.md

@@ -0,0 +1,30 @@
+---
+id: 2026-03-31T14:53:42-04:00
+aliases: []
+title: 2026-03-31 14:53:42
+tags:
+  - authorship/original
+  - destiny/permanent
+  - status/draft
+  - type/periodic/timestamped
+dg-publish: true
+date-created: 2026-03-31T14:53:42-04:00
+daily: "[[2026-03-31]]"
+weekly: "[[2026-W14]]"
+monthly: "[[2026-03]]"
+quarterly: "[[2026-Q1]]"
+yearly: "[[2026]]"
+---
+# 2026-03-31 14:53:42
+
+%% [[statistical-modeling-for-construction-estimating]] %%
+
+A while ago I wrote proofs
+for average distance between random points
+under taxicab geometry.
+
+I ought to find those and add them to the vault.
+
+One condition I know I didn't examine
+was random points on the perimeter of a rectangular space.
+This would be ideal for runs between equipment in the same room.