OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

# office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: [OASIS Issue Tracker] (OFFICE-3936) logarithmic scale misses information about basis

• From: OASIS Issues Tracker <workgroup_mailer@lists.oasis-open.org>
• To: office@lists.oasis-open.org
• Date: Sun, 29 Oct 2017 18:15:54 +0000 (UTC)

```     [ https://issues.oasis-open.org/browse/OFFICE-3936?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Andreas Guelzow updated OFFICE-3936:
------------------------------------

Proposal:
20.xx The chart:major-origin attribute specifies the location of one major tick.

20.xx The chart:minor-logarithmic attribute specifies whether the minor ticks are spaced equally after the logarithmic transformation. If this value is false they are spaced equally before the transformation.

20.27 The chart:interval-major attribute specifies major intervals on an axis 11.8.
If the axis is linear and the value of this attribute is x with the value of chart:major-origin being w, major ticks are placed at w+n*x for n being any (positive, negative or zero) integer.
If the axis is logarithmic and the value of this attribute is x with the value of chart:major-origin being w, major ticks are placed at w*(10^(n*x)) for n being any (positive, negative or zero) integer. (Note: if x = log_{10}(y), then 10^(n*x) = y^n.)

20.28 The chart:interval-minor-divisor attribute specifies a divisor for the chart:interval-major value, the division of which determines the minor interval:

If chart:minor-logarithmic is false the ticks are placed as follows:
If a and b are the values of consecutive major ticks (see 20.27) and m is the value of chart:interval-minor-divisor then the minor ticks are placed at locations corresponding to the values a + (b-a)/m * k with k = 1,2, ..., (m-1). So the minor ticks are equally spaced before the logarithmic transformation.

If chart:minor-logarithmic is true the ticks are placed as follows:
If a and b are the values of consecutive major ticks (see 20.27) and m is the value of chart:interval-minor-divisor then the minor ticks are placed at locations corresponding to the values a*((b/a)^(k/m)) with k = 1,2, ..., (m-1). So the minor ticks are equally spaced after the logarithmic transformation.

Schema Change:

--- OpenDocument-v1.2-os-schema.rng	2017-10-27 18:22:13.973894000 -0600
+++ OpenDocument-v1.2-os-schema+log-changes.rng	2017-10-29 12:12:35.544258788 -0600
@@ -17476,11 +17476,21 @@
</attribute>
</optional>
<optional>
+				<attribute name="chart:major-origin">
+					<ref name="double"/>
+				</attribute>
+			</optional>
+			<optional>
<attribute name="chart:interval-major">
<ref name="double"/>
</attribute>
</optional>
<optional>
+				<attribute name="chart:minor-logarithmic">
+					<ref name="boolean"/>
+				</attribute>
+			</optional>
+			<optional>
<attribute name="chart:interval-minor-divisor">
<ref name="positiveInteger"/>
</attribute>

was:
20.27
Add "for details see 20.28." to the first sentence. The wording would then be
The chart:interval-major attribute specifies major intervals on an axis 11.8, for details see 20.28.

20.28
Add the following text before the last sentence.

Let a be the value at a major grid line and b the value at the next major grid line and m the value of the attribute chart:interval-major.
For an axis with proportional scale it is m = b - a, for an axis with logarithmic scale it is m = log_10(b) - log_10(a) = log_10(b over a).
The distance D, measured e.g. in centimeter, between a pair of two adjacent major grid lines is the same for all such pairs.

Let n be the value of the attribute chart:interval-minor-divisor, and x_k the value at the k-th minor grid line with a < x_k < b.

Then for both kind of scale it is x_k = a + {{b - a} over n} cdot k .
But the distance d, measured e.g. in centimeter, from the major grid line at value a to the k-th minor grid is different.
For an axis with proportional scale it is
d = D cdot {k over n} = {D over m} cdot (x_k-a)
For an axis with logarithmic scale it is
d = {D over m} cdot (log_10(x_k) - log_10(a))

> logarithmic scale misses information about basis
> ------------------------------------------------
>
>                 Key: OFFICE-3936
>                 URL: https://issues.oasis-open.org/browse/OFFICE-3936
>             Project: OASIS Open Document Format for Office Applications (OpenDocument) TC
>          Issue Type: Improvement
>          Components: Chart
>    Affects Versions: ODF 1.2
>            Reporter: Regina Henschel
>            Assignee: Andreas Guelzow
>             Fix For: ODF 1.3
>
>
> http://docs.oasis-open.org/office/v1.2/os/OpenDocument-v1.2-os-part1.html#property-chart_interval-major
> Although the basis 10 is widely used, others a common too, e.g basis 2 for audiogram.
> This proposal adds not only the basis but additional details to the description of chart:interval-major and chart:interval-minor-division.
> Specifications are not only read by developers but by authors of guides and tutorials too. For those the mathematic impact of the attributes values might be not obvious.
> Formulas in the proposal are written StarMath.
> I think, that logarithmic scale should only be allowed for a value-axis, but have not included that in this proposal.

--
This message was sent by Atlassian JIRA
(v6.2.2#6258)
```

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]