Information Technology Reference
In-Depth Information
time of each point. Our implementation draws the segments of “Clear”
polylines in red, “Sunny” in yellow, “Cloudy” in green, “Rainy” in blue,
and “Snowy” in cyan, interpolating the colours if tags of two ends of a
segment are different. Exceptionally, it may draw the segments in grey if
we cannot obtain weather data. Using click or sketch interfaces, our
implementation draws selected polylines brightly and others in grey.
Observation of long-term trends
We visualized a dataset comprising temperature values from 1 to 31 July,
2009. Fig. 9.5 shows an overview of the temperature data. Fig. 9.5 (Upper)
shows an overview without tags, and Fig. 9.5 (Centre) shows an overview
with tags. Both show results, including all observation points, before
applying level-of-detail control. While Fig. 9.5 (Upper) displays just major
and outlier variations of temperature, Fig. 9.5 (Centre) displays much
more information. Colours in the dense regions denote major daily
weather. Blue (Rainy) is observed in several consecutive days and it
appears several times during the period. This result demonstrates the
effectiveness of the visualization of tagged time-varying data, since it is
difficult to obtain such knowledge from Fig. 9.5 (Upper). To explore
variation in greater detail, we applied level-of-detail control to the original
polylines including all observation points. Fig. 9.5 (Lower) shows the
results. In this figure, severe overlaps of polylines are reduced, especially
non-tagged ones, coloured in grey. We observe that rain persisted in
various regions. Also, we find only a few red or yellow (Clear or Sunny)
days, while there are about 10 to 20 sunny days in an average year as
shown in Table 9.1. This result demonstrates that this technique is suitable
for observing the long term time variation (e.g. one month) of values and
tags.
Search WWH ::
Custom Search