PALanalysis User Guide

PAL data files

.datx files

The datx file is the raw unprocessed data file from the activPAL. To generate the analysis outcomes, files need to be processed in PALanalysis (or by batch processing files in PALbatch).

PAL datx files can be opened in PALanalysis by:

  • Clicking browse to open a dialog box to locate the file
  • Dragging and dropping the datx file on to the window

(Windows version only) Double clicking on a datx file in Windows explorer will directly open the file in PALanalysis

.pml files

When opening datx files, PALanalysis will generate an accompanying .pml file. This stores user definitions, e.g. user defined Time in Bed. To retain any user defined changes, the pml file should be retained with the datx file.

Views

Clicking on the Views button opens the view selection menu. The different data views are outlined below.

Events

The events view shows the classified events on a heat map, with accompanying summary information.

The events can also be viewed on a continuous spiral, where each loop displays a day of data – the outer loop is the start of the recording, and the data spirals inwards to the centre.

The icons show the colour key for the heatmap events. The icons are clickable, to turn on and off the colours for individual classifications (excluding non-upright and upright classifications).

Highlight options

The “Highlight Upright” option varies the thickness of the heat map segments by event type, providing a greater visual separation of standing and stepping events from non-upright events.

The “Highlight Locus” option shows the classified locus information in the background of the heatmap. The light blue areas are the identified primary locus, and the light green areas are the identified secondary loci.

Summary data

The summary data shown can be changed in the sidebar options.

The sedentary bouts option highlights the duration of sedentary bouts in the data on a colour scale from yellow through to dark orange.

Locus filter

The heatmap can apply a locus filter which slices the heatmap and summary data using the primary locus, secondary loci or transitions between loci.

Events (hour per row)

This provides a detailed view of each individual day. The day displayed can be changed in the top bar menu. This view can be exported to pdf (Events linear (hour per row) pdf)

Epochs summary

This visualisation displays for each day an hour by hour epoch summary and daily summary. This view can be exported to pdf (Epochs (day per row) pdf)

File Details

This view provides some diagnostic information from the datx file.

Agnostic Analysis

This view displays a sensor location agnostic version of the analysis based on device inclination and acceleration. The data can be viewed linearly per day or on a spiral. The colours in the view below represent the 6 “dice face” primary orientations of the device, as used in the validation algorithms (MORA).

In the left side bar, the postural thigh-worn colour definitions can also be selected for comparison.

The acceleration sensor data can be overlaid on the inclination colours for context.

The acceleration magnitude can be visualised using the block height options, enabling the most dynamic sections of the data to be highlighted.

Alternatively, the block height can show the periods of stillness in the data – the higher the block, the longer the period of stillness in that orientation.

Accelerometer calibration

This view shows the accelerometer pre and post calibration points, in addition to statistics on calibration success. Note: this view is not relevant for the VANE and CREA algorithms which do not peform the calibration step.

Time in Bed adjustment

This visualisation shows the algorithm defined time in bed definitions, and allows for these to be adjusted.

The height of the height map bars indicates the stillness of the thigh, which can be used to correct the start and end times. These times can be adjusted by either dragging the start and end arrows, or entering specific times for each row.

Any changes are only applied by pressing the “Save Changes” button. The changes are saved to a “.pml” file, which is found in the same folder as the datx file (note: the original .datx file is not changed). To retain these manual changes, the pml file must be kept in the same folder as the datx file.

Sensor data

This view allows the accelerometer data to be inspected. The controls at the bottom of the screen enable zoom, pan and normalising the vertical axis to fit the data to the screen. Clicking on the menu button on the vertical axis provides options to adjust the data being shown.

A specific section of the data can be selected by clicking and dragging the mouse (indicated by the dark grey selection area).

Up to three signals can be shown on screen at once.

Exports

Export Selection

Enables the selection of the export options for batch processing.

The exports are grouped into different types: These are described on the following pages:

The “Include analysis parameters header” checkbox enables an optional header to csv exports to record the processing parameters which were used (Spreadsheet Header).

For convenience of tracking exports and the processing parameters a timestamped log file is created for each export:

/Documents/PAL/userlogs/PALanalysisExportLog_YYYYMMDD

Settings

The settings page is accessed by the cog icon on the right of the top bar. The following settings can be adjusted.

validation algorithm

Adjustable criteria for the validation algorithm. See algorithm desciptions (Validation algorithms)

Wear time protocol:

  • 24 hour wear – allows 4 hours of non-wear per day
  • 14 hour waking wear day – allow 10 hours of non-wear per day
  • 10 hour wear day – allow 14 hours of non-wear per day

valid day filter

Options for using the valid day flag as generated by the validation algorithm

  • Show all days – the valid day flag is not used
  • Highlight valid days – valid and invalid days are marked
  • Show only valid days – invalid days are trimmed from the start and end of the data

wear correction

With ‘Auto-correct inverted wear’ selected, data files where the device was worn upside down will be automatically corrected.

classification algorithm

The classification algorithm can be selected and parameters adjusted.
See algorithm descriptions (Classification algorithms).

Related Articles