Data Preparation v3 - RodentDataAnalytics/mwm-ml-gen GitHub Wiki

Contents

  1. Setting up the data from Ethovision
  2. The exported data format

Setting up the data from Ethovision

  1. Open Ethovision and select File-->Open Workspace.

    data_input_1

  2. On the dialogue box navigate to the folder containing the Ethovision's EWS file and open it.

    data_input_2

  3. Each Workspace file differ. The one used for this example contains data from three separate days, learning day 1 to learning day 3 (with bold letters the selected 'active' dataset is indicated).

    data_input_3

  4. To export the data first press the button 'Analyze Data' (the last button of the main tab).

    data_input_4

  5. Afterwards select File-->Export-->Parameters & Raw Data.

    data_input_5

  6. A dialogue box will appear. Fill it as below:

    • File Prefix: press 'Browse...' and navigate to a folder were you want the data to be extracted then give an appropriate prefix name for example 'day1'.
    • Save as Type: select 'Comma Separated Values (*.csv)'.
    • Include: tick all the checkboxes

    data_input_6

Important note: Later versions of Ethovision may not have the 'Comma Separated Values (*.csv)' option. In this case the option 'Excel' should be selected. This option will export the data in XLXS format (*.xlxs).

  1. Repeat steps 5 and 6 for the rest of the data but each time provide a different prefix name to avoid overwriting your previous exported data. Note, in order to select the next dataset for exporting right-click on its name and select 'Set as Active'.

  2. Assuming that there are three datasets and the selected prefixes were 'day1', 'day2' and 'day3' the folder selected for exporting should contain CSV files with the prefix 'day1_', 'day2_' and 'day3_' followed by a unique numeric id.

    data_input_7

The exported data format

The CSV (or XLXS) data format differ between the experiments but each exported Workspace keeps the same format over all the exported files. In general each set of exported files should contain the following:

  • a cell indicating the animal id followed by a cell which contains a numeric value,
  • a series of cells recorded time, x, y. Cells that comes below them indicate the location of the animal over time.

data_format_1

The cells indicating the animal id, recorded time, x and y will be required later when a new project is being created.