bpy Vectors comparison #43622

New Issue

Giuseppe De Marco · 2015-02-10T20:57:06+01:00

Giuseppe De Marco commented

2015-02-10 20:57:06 +01:00

System Information
Debian 8 - Jessie

Blender Version
bpy.app.build_hash: bbf09d9
bpy.app.version_string: '2.73 (sub 0)'

This isn't properly a bug, but I cannot understand what's happening here:

Writing a simple addon used for increase / decrease and set an
arbitrary edge length I'm learning the way that blender uses to work
with vertices and edges. I noticed this, in python console:

from mathutils import Vector

v3 = Vector((1.0, 0.0, 0.0))
v6 = Vector((-1.0, 0.0, 0.0))

v6 == v3
False

v6 > v3

False

v6 >= v3

True

Who could explain what's is the logical approach to this ?
Is there some arithmetic initial conditions that would be usefull to know before working with vertices ?

**System Information** Debian 8 - Jessie **Blender Version** bpy.app.build_hash: bbf09d9 bpy.app.version_string: '2.73 (sub 0)' This isn't properly a bug, but I cannot understand what's happening here: Writing a simple addon used for increase / decrease and set an arbitrary edge length I'm learning the way that blender uses to work with vertices and edges. I noticed this, in python console: ``` from mathutils import Vector v3 = Vector((1.0, 0.0, 0.0)) v6 = Vector((-1.0, 0.0, 0.0)) ``` v6 == v3 False >>> v6 > v3 False >>> v6 >= v3 True Who could explain what's is the logical approach to this ? Is there some arithmetic initial conditions that would be usefull to know before working with vertices ?

Giuseppe De Marco commented

2015-02-10 20:57:06 +01:00

Changed status to: 'Open'

Giuseppe De Marco commented

2015-02-10 20:57:06 +01:00

Added subscriber: @demarcog83

Sybren A. Stüvel self-assigned this 2015-02-10 21:14:22 +01:00

Sybren A. Stüvel commented

2015-02-10 21:21:12 +01:00

Changed status from 'Open' to: 'Resolved'

Sybren A. Stüvel closed this issue

2015-02-10 21:21:12 +01:00

Sybren A. Stüvel commented

2015-02-10 21:21:12 +01:00

Vector comparison is done in two ways. The == and != operators perform equality checks as you'd expect: v3 == v6 implies that all components are equal. However, component-wise ordering doesn't make sense, and <, >, <= and >= compare vector length instead. The result is that v3 > v6 or v3 == v6 is not the same as v3 >= v6.

Vector comparison is done in two ways. The `==` and `!=` operators perform equality checks as you'd expect: `v3 == v6` implies that all components are equal. However, component-wise ordering doesn't make sense, and `<`, `>`, `<=` and `>=` compare vector *length* instead. The result is that `v3 > v6 or v3 == v6` is not the same as `v3 >= v6`.

Giuseppe De Marco commented

2015-02-11 09:46:00 +01:00

Thank you sybren,

This situation doesn't seems so rationale, I perfectly understood what you mean however there're many imperfections that could discourage users in deepen blender.

"""
The result is that v3 > v6 or v3 == v6 is not the same as v3 >= v6.
"""

In this case it should return always false, isn't so ?

For the addon I'm developing, mesh_edges_length.py, I start coding functions to do comparison of vectors in absolute values and other function to do vector comparison. Don't you think that these features should be configured as "primitives" in blender and not extended through python addons ?

Thank you sybren, This situation doesn't seems so rationale, I perfectly understood what you mean however there're many imperfections that could discourage users in deepen blender. """ The result is that v3 > v6 or v3 == v6 is not the same as v3 >= v6. """ In this case it should return always false, isn't so ? For the addon I'm developing, mesh_edges_length.py, I start coding functions to do comparison of vectors in absolute values and other function to do vector comparison. Don't you think that these features should be configured as "primitives" in blender and not extended through python addons ?

Sybren A. Stüvel commented

2015-02-11 11:22:42 +01:00

The absolute value of a vector is its length, so that's already in Blender. Implementing component-wise inequality (like >= per component) would need to return a Vector such that (a, b) >= (c, d) --> (a >= c, b >= d).

At this moment it's still quite an abstract discussion. Personally I wouldn't need this feature, as it goes against what I know of vectors (the mathematical concept, not the Blender class in particular). However, we welcome all improvements. If you have code that implements such features, and a good demonstration of why people need this, of course we'll look at it.

The absolute value of a vector *is* its length, so that's already in Blender. Implementing component-wise inequality (like `>=` per component) would need to return a Vector such that `(a, b) >= (c, d) --> (a >= c, b >= d)`. At this moment it's still quite an abstract discussion. Personally I wouldn't need this feature, as it goes against what I know of vectors (the mathematical concept, not the Blender class in particular). However, we welcome all improvements. If you have code that implements such features, and a good demonstration of why people need this, of course we'll look at it.

Giuseppe De Marco commented

2015-02-11 11:24:26 +01:00

Thank you sybren

Giuseppe De Marco commented

2015-02-12 14:28:50 +01:00

from bf-python ML :

Date: Tue, 10 Feb 2015 21:41:02 +0100
From: Sybren A. St?vel sybren@XxXXxXXXxl.eu
Subject: Re: [Bf-python] Vectors comparison doubts
To: Blender Foundation Python list bf-python@blender.org
Message-ID: 2685462.Pa2RXgPjun@herrie
Content-Type: text/plain; charset="utf-8"

On Tuesday 10 February 2015 09:22:56 Mark Young wrote:

?Unless something's changed since 2.49, when you do an equality check for
vectors, it does an element-wise comparison. When you use the inequality
operators, it does a magnitude comparison.

That's also what's currently happening.

Guiseppe's email did make me check the documentation, and I noticed a typo in
there, so that'll be fixed in the next Blender release too :)

from bf-python ML : Date: Tue, 10 Feb 2015 21:41:02 +0100 From: Sybren A. St?vel <sybren@XxXXxXXXxl.eu> Subject: Re: [Bf-python] Vectors comparison doubts To: Blender Foundation Python list <bf-python@blender.org> Message-ID: <2685462.Pa2RXgPjun@herrie> Content-Type: text/plain; charset="utf-8" On Tuesday 10 February 2015 09:22:56 Mark Young wrote: > ?Unless something's changed since 2.49, when you do an equality check for > vectors, it does an element-wise comparison. When you use the inequality > operators, it does a magnitude comparison. That's also what's currently happening. Guiseppe's email did make me check the documentation, and I noticed a typo in there, so that'll be fixed in the next Blender release too :)

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

bpy Vectors comparison #43622