A Tool for Network Visualization
A csv file containing the information of nodes. Columns 'Id' and 'Label' are required.
A csv file containing the information of edges. Columns 'Source' and 'Target' are required, which depend on the 'Id' column of the node file.
All of the nodes are of the same size.
sizeValue: The size value of all nodes.
Sizes are ranked by certain attribute of nodes.
Method: The attribute which the nodes' sizes depend on. The supporting attributes include the in-degree(number of nodes point to the node), out-degree(number of nodes the node points to), degree(sum of in-degree and out-degree) and other numerical attributes of the nodes contained in the csv file.
minSize: The minimum size of all nodes.
maxSize: The maximum size of all nodes.
All of the nodes are of the same color.
colorValue: The color value of all nodes.
Nodes are partitioned into different groups according to their attributes and nodes within a group are of the same color.
Method: Nodes are partitioned into different groups according to the attribute and nodes within a group are of the same color.
Colors are ranked by certain attribute of nodes.
Method: The attribute which the nodes' colors depend on. The supporting attributes include the in-degree(number of nodes point to the node), out-degree(number of nodes the node points to), degree(sum of in-degree and out-degree) and other numerical attributes of the nodes contained in the csv file.
minColor: The minimum value of color of all nodes.
maxColor: The maximum value of color of all nodes.
All of the edges are of the same color.
colorValue: The color value of all edges.
Colors are ranked by certain attribute of edges.
Method: The colors of edges are ranked by the weights.
minColor: The minimum value of color of all edges.
maxColor: The maximum value of color of all edges.
Quality layout: A linear-attraction linear-repulsion model with few approximations. Speed automatically calculated.
gravity: Attract the nodes to the center. Prevents islands from drifting away.
strongGravityMode: A stronger gravity law.
barnesHutTheta:Theta of the Barnes Hut optimization.
edgeWeightInfluence:How much influence you give to the edges weight.
scalingRatio:How much repulsion you want. More makes a more sparse graph.
barnesHutOptimize:Barnes Hut optimization.
adjustSizes:Use only when spatialized.
Original Yifan Hu's attraction-repulsion model. Reduce the computational cost by restricting force calculation to the neighborhood. The algorithm stops itself, as it has an alternative cooling sheme.
barnesHutTheta: The theta parameter for Barnes-Hut opening criteria. Smaller values mean more accuracy.
quadTreeMaxLevel: The maximum value to be used in the quadtree representation. Greater values mean more accuarcy.
initialStep: The initial step size used in the integration phase. Set this value to a meaningful size compared to the optimal distance.
convergenceThreshold: Relative energy convergence threshold. Smaller values mean more accuracy.
optimalDistance: The natural length of the springs. Bigger value mean nodes will be farther apart.
relativeStrength: The relative strength between electrical force(repulsion) and spring force(attraction).
adaptiveCooling: Controls the use of adaptive cooling. It is used to help the layout algorithm to avoid energy local minimal.
stepRatio: The ratio used to update the step size across iterations.
ForceAtlas makes graph more compact, readable, and can show authorities more central than hubs. Auto-stabilization improve convergence at the end of the layout.
inertia: Conservation of node speed at each new pass.
repulsionStrength: How strong does each node rejects others.
maxDisplacement: This is limiting each node's displacement.
speed: Permits to increase convergence speed at the price of a precision loss.
freezeStrength: Power of the auto stabilize function.
f1gravity: This force attracts all the nodes to the center to avoid dispersion of disconnected components.
freezeInertia: This parameter represents the auto-adaptiveness of the anti-flickering function.x
freezeBalance: Activates the freezing of unstable nodes. Despite some loss of efficiency, this prevents most of nodes flickering.
Fruchterman Reingold is a classical layout function, since 1984.
gravity: This force attracts all nodes to the center to avoid dispersion of disconnected components.
speed: Increase convergence speed at the price of a precision lose.
area: The graph size area.
Label Adjust works on text size to repulse nodes and therefore makes every label readable.
speed: Speed factor.
adjustBySize: Include node size in repulsion.
Label Adjust works on text size to repulse nodes and therefore makes every label readable.
speed: Speed.
ratio: Ratio.
margin: Margin.
Rotate the graph by degrees.
angle: Clockwise rotation angle in degrees.
Contract the layout around its center.
scale: Scale factor.
Expand the layout away from its center.
scale: Scale factor.
A random distribution of the nodes.
size: The size of the space to randomly distribute the nodes.
The positions of nodes are pre-defined.
position: The csv file which contains 'x' and 'y' columns.
Developed by the Lab for Big Data Methodology