Graph Settings

This page describes the form controls available on the Fitbit page that allow you to customise the appearance of the graph thereon.

General

Allow zoom on vertical axis

Dragging with your mouse will zoom in on the vertical (value) axis (as well as the timescale axis).

Note: Due to a bug in the graphing library, deselecting this option can result in ugly (non-round) values being used on the horizontal (time) scale.

Raw Data

Draw line

Draws a blue line on the graph that represents the raw data from Fitbit.

Max. sequence interval

The maximum number of seconds between values that will be connected with a line. When there are no values for longer than this (eg, due to the device not being worn or worn correctly), there will be a gap in the line.

Smoothed Data

Draw line

Draws a green line on the graph that smoothes out short-term variations in the raw data.

Max. sequence interval

The maximum number of seconds between raw values that will be considered to form a contiguous set. Smoothing will not occur between such sequences; ie, each sequence will be treated independently of the others.

Time step

The number of seconds between successive smoothed values in the same sequence.

Window width

The number of seconds from which raw values will be used to calculate each smoothed data value. For example, if the window width is 60 seconds, data values up to 30 seconds either side of each point on the smoothed line will be taken into account (excepting values that are not considered to be part of the same sequence).

Interpolate

Controls whether raw data will be interpolated between values, for the purpose of computing smoothed values. The raw data usually comprises readings that are a few seconds apart. When interpolation is used, values for each second in between those readings are computed by interpolating the readings to either side. When interpolation is not used, values for each second in between readings are assumed to be the same as the nearest reading.

Centre-weight

Controls whether data values near the centre of the window will be weighted more highly than those further away.

When centre-weighting is used, a triangular weighting algorithm is employed such that values at the edges of the window (ie, half of the window width away from the time point being considered) will be given zero weight, while the value at the centre of the window (ie, at the exact time point being considered) will be given maximum weight. Intermediate values will be weighted in proportion to their distance from the time point being considered.

When centre-weighting is not used, all values within the window are weighted equally.