DataTube2 User Guide

This site gives you the initial and necessary information to run DataTube2, a visual mining application devoted to temporal data.

Outils pour utilisateurs

Outils du site


loading

Ceci est une ancienne révision du document !


Loading demo files or your own data

Synchronous and asynchonous temporal formats

DataTube2 considers its own two CSV formats for temporal data:

  • The synchronous temporal format is when all data attributes (i.e., variables) were recorded at the same time steps and with the same temporal difference between time steps. Hence, it is a matrix, like in this example:
Time step Variable1 Variable2
2003-09-28 00:00:00 10 15
2003-09-29 00:00:00 12 15
2003-09-30 00:00:00 14 12

If you use the synchronous format, be sure that the two above conditions are met.

  • The asynchronous temporal format considers that data attributes were recorded at any time and independently. This is an “event” type representation, like in this example:
Time step Variable name Value
2003-09-28 00:00:00 Variable1 10
2003-09-29 00:30:00 Variable1 15
2003-10-01 00:00:00 Variable1 13
2003-09-29 01:30:00 Variable2 15
2003-11-01 00:00:00 Variable1 13

This array does not need to be sorted.

CSV format for synchronous files: French (separator = ";", decimal =",") or International (separator = ",", decimal = ".")

To begin, let us remind you that the differences between the French CSV format and the International CSV format are the following:

French CSV format International CSV format
Cell separator ; ,
Decimal separator , .

The DataTube2 synchronous format supports both encodings, and you can specify the French/International encoding by answering “Yes” or “No” in the following window, which is diplayed when you open a file:

CSV format for asynchronous files: International (separator = "TAB", decimal = ".")

The asynchronous format in DataTube2 uses TAB character as separator and “.” for decimals:

Loading and visualizing the demo files

Simple demo files with artificial data are provided in each format (synchronous or asynchronous) and for each CSV encoding (French or international).

DT2-synchronous-small-file-international.csv

Encoding your own data

To encore your own data in a CSV file:

  • Select the format to use, i.e., either synchronous or asynchronous (see explanation above),
  • Study the examples provided with DataTube2 (see the Data directory),
  • Give names to the variables,
  • To represent time values :
    • the GMT format is yyyy-mm-dd hh:mm:ss
    • yyyy must be greater than 1970 (this is due to the EPOCH format, we'll try to improve in future releases),
    • DataTube2 can use GMT or EPOCH (but yet, please use GMT, we need additional testing for EPOCH format),
  • To represent data values:
    • for synchronous format (see the discussion above), use either the French CSV format (like this “;3,1415;”) or the International CSV format (like this “,3.1415,”),
    • for asynchronous format, use TAB for separators and “.” for decimals,
loading.1424354185.txt.gz · Dernière modification: 2015/02/19 14:56 (modification externe)