China
Africa
Explore chapters and articles related to this topic
Structured Estimation in High Dimensions
Published in Ashok N. Srivastava, Ramakrishna Nemani, Karsten Steinhaeuser, Large-Scale Machine Learning in the Earth Sciences, 2017
André R Goncalves, Arindam Banerjee, Vidyashankar Sivakumar, Soumyadeep Chatterjee
Figure 2.8 presents the dependency structure using a chord diagram. Each point on the periphery of the circle is a location in South America and represents the task of learning to predict temperature at that location. The locations are arranged serially on the periphery according to the respective countries. We immediately observe that the locations in Brazil are heavily connected to parts of Peru, Colombia, and parts of Bolivia. These connections are interesting as these parts of South America comprise the Amazon rainforest. We also observe that locations within Chile and Argentina are less densely connected to other parts of South America. A possible explanation could be that while Chile, which includes the Atacama Desert is a dry region located to the west of the Andes, Argentina, especially the southern part experiences heavy snowfall which is different from the hot and humid rain forests or the dry and arid deserts on the west coast. Both these regions experience climatic conditions which are disparate from the northern rain forests and from each other. The task dependencies estimated from the data reflect this disparity.
Visualizing data
Published in Celia Lury, Rachel Fensham, Alexandra Heller-Nicholas, Sybille Lammes, Angela Last, Mike Michael, Emma Uprichard, Routledge Handbook of Interdisciplinary Research Methods, 2018
Design Space might initially appear to be small for simple data sets. For instance, a scatterplot might seem the only choice when visualizing two vectors of continuous data, such as for a category in Anscombe’s Quartet (e.g. XI, YI in Figure 2.12.2). Yet the axes of a scatterplot can be aligned to produce a parallel-coordinates plot, then bent to simulate a chord diagram or hive plot, or the values can be summed for a stacked bar chart, which can be bowed into a pie chart and then punctured to produce a donut plot. Each jump in Design Space can modify the meaning and information content (Figures 2.12.3, 2.12.4 and 2.12.5), even when the symbols, shapes and scales are unchanged.
Using Pro Tools
Published in Mike Collins, Pro Tools 9, 2012
The Chords ruler lets you add, change, move, and delete chord symbols. To add a chord symbol, place the cursor in the Timeline where you want to add this. Click the Plus (+) button in the Chords ruler to open the Chord Change dialog at the current Timeline location. In the Chord Change dialog, you can select the name for the root of the chord, the chord type, the bass note of the chord, and the chord diagram that will be displayed in the Score Editor.
Story Analysis Using Natural Language Processing and Interactive Dashboards
Published in Journal of Computer Information Systems, 2022
The lists of people and groups at the left of this section are paired with their impact within the narrative. This impact measures not only the number of mentions within the narrative, but also the number of actions and impacting subjects or impacted objects for the person or group. Higher impact actors are at the top. The circular Interactions visualization is a chord diagram. The bands along the circumference represent people (blue) or groups (red) in the story, and the connections (chords) between these bands represent the interactions between them. These relationships can occur between two people, two groups, or a person and a group. The final visualization in this section is the Narrative Web. This is an example of a force-directed graph. Nodes in this web represent people, organizations, countries, and other entities, color coded for visual clarity. Hover over a node and a tooltip will appear containing the word and entity type. The links between nodes represent interactions between these entities. Generally, the bulk of the action will be clearly visible in a cluster of nodes in the center of the web.
Can we design an industry classification system that reflects industry architecture?
Published in Journal of Enterprise Transformation, 2018
Margaret Dalziel, Xiangyang Yang, Simon Breslav, Azam Khan, Jianxi Luo
We now turn to our second proposition on the nature of interindustry relations within demand-based vertical sectors. We expect that when industries within demand-based vertical sectors are classified according to their role, interindustry relations will be hierarchically structured, with firms in industries in upstream subsectors generally selling to firms in industries in downstream sectors. The chord diagrams in Figures 5–8 depict the transactions that originate or terminate in the four sectors that are best represented in our dataset: transportation, information and communications technologies (ICT), health, and energy. Collectively, sales from sellers in these four sectors represent 81% of the value of transactions in our dataset. A chord diagram arranges the nodes radially, drawing thick chords between nodes. To best show within-sector relationships between roles, sector roles are shown in horizontal text on the outer ring, and sectors are shown in radial text on the inner ring, occupying areas that are proportional to their representation as sellers in the dataset. These proportions are constant across the four diagrams, sellers that are parts and materials suppliers, shown in black, dominate the outer ring, occupying 38% of the circumference, and sellers that are parts and materials suppliers in the transportation sector, shown in green, occupy approximately 18% of the circumference. Nodes are role-sector pairs and the thickness of the chords between nodes indicates the net value of transactions as a proportion of all transactions. Where a chord is tapered, there are more sales than purchases by the node at the thick end of the chord (Holten, 2006). By showing net transactions between nodes, rather than all transactions, chord diagrams make the patterns in the data visible.