Time-series MST
  • Data
    • Upload
    • Settings
    • Small demo data
    • Full demo with sectors
    • Full demo data without sectors
    • Printer signal demo
  • View
    • Single MST
    • Two MST
    • Evolving MST
    • Systematic analysis (demo only)
    • MST visualization
    • Graph visualization
    • Nodes represent symbols
    • Nodes represent classes
  • Help
  • Update graphs
Required Format
  • First row is for symbols
  • First column is for timestamp
  • Fields are separated with either comma or tab
  • NaN values should be represented with NaN or \N
  • Classes for the symbols can be put on the second row and their id must be named 'classes'
  • Preprocessing
  • Filtering



Check your interested symbols
Paste ticker names in one column or separate them with commas (empty field means all tickers)
  • stock day on a daily, weekly or monthly basis.
  • calculate the value difference between adjacent time points.
  • the lowest admissible correlation value for edges.
  • percentage of available time points above which a symbol is qualified.
  • are time series elements, a.k.a. signals, tickers(stock market).
It is suggested to set a high value in order to prune unnecessary edges. You can change the value in the setting menu later on.

Graph A for time period: Value


Graph B for time period: Value


Statistics for graph A

Number of edges:
Edge weights:
Radius:
Modularity score:

Degree distribution

Statistics for graph B

Number of edges:
Edge weights:
Radius:
Modularity score:

Evolving correlation graph for time period: Value


Window controller

Statistics

No. edges:
Edge weights:
Radius:
Modularity score:

Degree distribution

Current file:
Class labels:
Debug Area:



Acknowledgment

This work is sponsored by the ISMARD project which is a joint project of the Dept. of Computer science and the Dept. of Economics at the Free University of Bozen-Bolzano

Institution



Useful links

Reference

Help

Notebook version

Contact

Piazza Domenicani 3, Bozen, Italy

XXXX@XXX.XXX

+ 39 XXXXXXXXXX