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.
Q7 Answers
- Is this assumption correct? In small multiples the data is split into different item sets and in multiviews different attributes of the whole data set are shown.
-
- How to decide which view to put where on the screen? Or let the user rearrange them? When to use different windows an when is it better just to create regions inside one window?
-
- Is clustering a form of partitioning and grouping? How to find a good way for grouping?
-
- Are there tools, where the user decides how to group and which group to show where and how?
-
- Is there a danger that information might not be visible when using microglyphs, e.g. as there are no borders?
-
- What is a general approach to find good layers and good partitions? Or how to decide which views/glyphs to use?
-
- Ich habe eine Frage zu den partitioning attributes. Partitioning kann also mit key oder value attributes realisiert werden. Aber wie kann ein attribute caterogical sein, ohne ein key zu sein?
-
- Eine Frage zu superimpose layers: Was genau muss ich bei der letzten Designentscheidung wie man die einzelnen items auf den layers verteilt beachten? Mit dieser Erklärung kann ich leider nicht viel anfangen. Das Beispiel 12.14 ist dann klar, aber die Theorie dahinter wie gesagt für mich sehr undurchsichtig. Vielleicht nur weil es eine wissenschaftliche Lektüre ist :)?
-
- In diesem Kapitel wurde sehr lange versucht den Unterschied zwischen einem view und einer glyph zu erklären. Im Endeffekt ist es aber doch das selbe, je nachdem wie man es betrachtet. Für die Entwicklung eines Visualisierungstools in meinen Augen eigentlich irrelevant? Denn scheinbar gibt es keine relevanten Aspekte in der Theorie der Visualisierung, die explizit mit glyphs arbeiten.
-
- When would it be better to use the 3D-perspective? It seems like it has no real benefits…
-
- Figure 12.15. Which approach is better? I personally like a) better then b) because you can compare everything easily.
-
- Figure 12.17. Is this the same approach as we saw in the last course from the video? I don't really remember the name of it.
-
- Are there good resources that cover the computational cost of the visualization methods? I think that especially the recursive subviews could be quite expensive to create.
-
- While I already knew some methods as Matrix Alignments a lot of methods are totally new to me. Many of them seem to be relatively domain specific and I am not sure whether one could use the same machinery for 'general' datasets. Is there a good framework that allows you to try these facets with your own data? ( Maybe Tableau allows this, but I have to little experience with it ).
-
- How many multiple views are too much? Are there any problems linked with information overload? (see Figure 12.7. The Improvise toolkit looks too demanding.)
-
- Are there any differences regarding user interaction performance between multiple views for the same data type and multiple views for different data types?
-
- How do we determine the ideal number of layers to superimpose? How much is too much?
-
- In Figure 12.7 there are a lot of different views of the data. You can see, that there are a lot of filtering methods. It's quite hard to understand the whole visualization. What I'm thinking about: If I change one of those views by filtering for example the central scatterplot in the bottom by mark a specific window of data ( is it called lassoing? ), does the whole visualization change in that case?
-
- My next question is about partitioning of barcharts. Can you say that there is a rule where to use single barcharts with grouped bars or where to use aligned small multiple barcharts, which we can see in Figure 12.8.
-
- When creating a superimposed linechart can I assume that the axis dimensions are compatible or is it necessary to update those? But if they visualize the same data, why aren't the combined in the first place?
-
- 12.2.: partitioning and showing only specific parts of the data: When you use an interactive tool and you have the opportunity to give the user the possibility to choose every aspect of the data, how many possibilities are to much for an user to explore?
-
- 12.3.1.: in the first assignment i opted to show humidity and temperature side by side. The problem was that there were more data sets for temperature than humidity. I wanted to keep both time axis identical to make it easier to compare. Should i have left the missing data visible or limit the time i displayed?
-
- 12.4.: partitioning into views: In an interactive visualization should we give this opportunity to the user if possible, or is that to much respectively do we lose a channel to display data?
-
- At recursive subdivision, the user can customize the visual, isn't it counterproductive to let him/her customize it, especially in the sight of how we show the data?
-
- at map layering, are there standarizations for the blue of the ocean, yellow of land, … or is it just important that they have different luminance?
-
- Is the use of dynamic layers just at computers (interactive) possible or also makeable at print publications?
-
- How many juxtaposed views (like a map in a map…) are reasonable?
-
- What does HiVE mean?
-
- In Fig. 12.11 I don't see the conclusion of chart a) and b). What do we get so differently when we switch the splits' orders?
-
- What's the best choice to combine shared data with?
-
- Why isn't usual to combine shared data with shared encoding, if it's often used in overview-detail combined with navigation support?
-
- In the example "Multiform Overview-Detail Microarrays", the book says that the list view provides a simple overview and helps the user to quickly select a name, but is this view still efficient even if the list contains a lot of entries?
-
- In section "Juxtapose Views" it is mentioned that the vis designers sometimes leave the arrangement to the window system running on the user's computer. What's the best way to do so and is it possible to leave not only the arrangement of views but also another characteristics to the built-in functionality?
-
- Für welche Daten verwendet man "Magnification Fields" ?
-
- Wie kann man die Verzerrung messen, damit man es bekannt geben kann, wenn es nötig ist?
-
- Meiner Meinung nach kann alles, was in 3D gezeigt wird, kann auch in 2D gezeigt werden. Stimmt das?
-
- Glyphs are not repeatable symbols with distinctive meanings but more generally very small areas used to present data in a certain, for the individual glyph specific, way. Is this correct?
-
- In section 12.5.1 they say more then three layers is very hard to achieve, if every layer contains many area marks, but do not all maps actually have quite a lot of layers and a substantial amount of static superimposed layers? They also say having many layers is only possible if each layer contains very little, i.e. a single line, but figure 12.13 shows many layers with different types of marks combined and the image is still comprehensible?
-
- Since small multiples work very well with animations and animations have an inherit time-axis to them, the ordering of the small multiples has to be chronological, to not confuse the user. Is there a way of making comparisons easier between non-adjacent small-multiples, without changing their position?
-
- (Are parallel coordinates superimposed (static) layers of multiple line graphs?)
-
- On page 288 it says that superimposed layers are completely transparent where no marks exist. Wouldn’t it make sense in some cases to make the superimposed layer only semi-transparent, to highlight its content against the lower layers?
-
- Is there a standard approach to ordering the attributes within a grouped bar chart? The usual approach to order the bars by magnitude obviously does not work here, since for each group a different attribute could have the highest bar, but the order of the attributes should not change between groups.
-
- Are there any resources about text-views in VIS? I understand that the VIS community probably does not focus on text, but as seen on p273,274, text views can enhance VIS-systems.
-
Letzte Änderung: 07.05.2015, 18:44 | 1357 Worte