Please disable Adblockers and enable JavaScript for domain CEWebS.cs.univie.ac.at! We have NO ADS, but they may interfere with some of our course material.
Q6 Answers
- Why are matrix views not able to show topological structure?
-
- For some views users have to be trained to use them. What is the tradeoff between training a user to use a more complex view and simply providing views that are understandable without training?
-
- How to find colours that are "as different as possible" when using colour as a categorical channel in RGB space?
-
- Using hue as ordered channel with "learned" colour sets as traffic light should work. In the lesson it was said, that the colours have different meanings in different cultures. Are there studies on this topic? How to find out about such issues?
-
- "The effectiveness of the general idiom varies considerably depending on the specific visual encoding idiom used; there has been a great deal of algorithmic work in this area." (Node-Link, p. 210) Is there an example for these algorithms?
-
- What is shown in Figure 10.3? The Desciption does not really help.
-
- What is meant by "Motion channels"? Is she talking about the motion in Dynamic charts?
-
- The spline radial layout tree seems to make things more complicated.Even for a small number of nodes and connections.This is just a comment, and I would not try to argue that it has moredisadvantages than benefits.
-
- How important is cognitive load? How can we choose the right amount of cognitive load?
-
- What is a good diverging color map to show the difference between two scalar values?
-
- Tamara is saying that the RGB system is a "very poor match for the mechanics of how we see". I kind of don't get that statement. Let's say the RGB standard is a 24-bit depth. which means that that color space holds about 17million colors. And there are even RGB models with a larger amount of bits. Thus our eye has cones witch highly react on blue, green and red wavelength I didn't really understand why it's that poor. I mean I understand that the HSL/HSV-model is more accurate, because the gamut is build upon the possibility to change luminance and saturation after picking a hue-value. But is there a better explanation why the RGB system is that bad - apart from being bound to generate the color with the 3 basic colors?
-
- How can luminous intensity be measured in real life? Just wondering.
-
- Figure 10.6 is showing different cases of color encoding. I'm just wondering about the case in which a two-dimensional diverging data is shown. There the color encoding doesn't make that much sense. Or does it? The green color is a completely other color-value, despite the light-pink colors a quite similar. Am I wrong or is it a poor color encoding decision to visualize it like this?
-
- In figure 9.10c there is a hierarchical network visualized. I'm just wondering about the colors of the root nodes. All of the "end-ponts" are colored blue, but speaking of hierarchical data, they shouldn't be on the same level?
-
- Beside the compactness of the radial tree layout in Fig. 9.2 is there any difference in the two node-link trees a) and b)?
-
- In 9.4 the message is that matrix views are better in scalability than node trees… I cannot agree here because it is very hard to get a visualisation of a network with its connection by looking at a matrix. Sure, a node-tree needs ahigh amount of space when scaling, but I would only draw a matrix when I only need to look up a certain connection e.g. when I want to know what is connected to node n. Am I missing something?
-
- Why are there different color systems like S-RGB and HSL if they want to show the same?
-
- In which way can the layout algorithm's calculations provide the information density and guarantee that there are no edge crossings or node overlaps?
-
- Which one is a better choice to represent large networks, a matrix view or a treemap?
-
- What kind of effects does it have if we use transparency with luminance and saturation?
-
- Tamara is saying that luminance is bound to 2-4 bins, hue is bound to around six or seven bins. And in the lecture we heard that rainbow coloring is a no-go. Whats the best way to color encode a vis with 256 bin for example.
-
- "Ware suggests avoiding grayscale if more thant two to four bins are required". Does this mean, that I can use the luminance channel of a specific colour for ordered data types greater than four. In the book it's just said that you should avoid grayscales in that case.
-
- Can you say that a problem visualized with a node -link view get's more readable as a matrix view by increasing the nodes and links?
-
- Wenn der Graph nach der Erzeugung immer anders aussieht, kann dies dann zu einer anderen Interpretation führen, im Bezug auf, was mir als erstes ins Auge springt ist am wichtigsten?
-
- Ist bei den Netzwerken das einzelne Datum noch von Bedeutung oder geht es nur mehr um die Verbindung untereinander?
-
- Wirken Bewegungen in Diagrammen nicht iritierend oder störend, wenn zum Beispiel mein Punkt immer auf und ab hüpft?
-
- Warum kann Partitioning mehrere Views anbieten, aber Layering nicht?
-
- Ist HIVE ein Treemap? Oder benutzt es Treemaps?
-
- Was ist der Unterschied zwischen "dynamic layer" und "static layer" ?
-
- Figure 10.3: Why has L from HSL always the same shade of grey?
-
- How is it possible to color encode a larger dataset of categorialized data?
-
- How to build a vis of a topology with a lot of different nodes and a lot of hops?
-
- 3D layout can be dangerous but could it be useful in this case?
-
- Regarding matrix views: Are the cells of the matrix really area marks (I would interpret the text that way)? What is the difference between area- and point mark then? I mean both have a certain area, but area does not seem to be important for the equally-sized matrix cells, couldn’t they just as well be rectangular ‘points’.
-
- Why use low-saturated colors for large areas? I did not find the reason in the text.
-
- Why is stippling for 2D areas part of the texture channel and for lines within shape channel?
-
Letzte Änderung: 07.05.2015, 18:38 | 1045 Worte