pedarProbe.node#
Experimental data is usually stored in a multi-layer way, which cast server difficulty to effectively aggregate them and implement analysis. In a typical pedar data folder, data separate in various files can be identified as a 5 layer structure: subject - condition - trail - foot - stance.
To formulate a universal framework for data analysis, it’s an appealing choice to build a node trees that consist of these layers to store the data.
Note
Python dictionary is very convenient for creating node tree: every dictionary object is a node in the node tree, and its branch nodes are added as a value to the dictionary, with its name as the keyword, layer by layer. In this case, any node in the node tree can be called in the format: root[subject][condition][trail][foot][stance]. However, unlike other python object, it’s not convenient to add new attributes to it. Therefore a dictionary object is not capable to realise all required features of a node tree, in respect to the classical computer science’s view.
For this reason, various classes are developed supporting for construction of the data analysis workflow:
Nodeclass is derived from the dictionary classDictto realised the basic node’s features.DynamicNodeis derived fromNodeto realised the layer layout restructuring feature.PedarNodeis derived fromDynamicNodeconveying pedar data analysis and result aggregation through the whole node tree.DataNodeis derived fromPedarNodewhich is actually the leaf node of the node tree, storing the raw data prepared for analysis.
Tip
Such framework can be easily transferred for other data analysis task, especially the Node and DynamicNode which were developed in a highly generalisable way.
Classes
Derived from |
|
Derived from |
|
|
|
Derived from |

Inheritance Relationship#