4. jbsteps.csv and gfsteps.csv - LisaGotzian/HeartSteps GitHub Wiki

Using the Jawbone tracker as well as the steps measured using the phone accelerometer ("Google fit"), a participant's steps were recorded throughout the intervention. The minute-by-minute steps are recorded in jbsteps.csv and gfsteps.csv, a 197,524x8 dataframe for steps from the Google Fit information and a 237,865x11 dataframe for steps from the Jawbone tracker dataframe for each user for each minute.
We give you two possible identifiers for merging with dailyema:

  • decision.index: the current decision time slot, so steps walked since notification was sent
  • decision.index.following.dec: the following time slot, so steps walked before notification was sent

identifier: user.index, ema.index, decision.index and steps.utime

Columns Description
user.index User ID
ema.index Ema ID
steps.utime Date-time minute at which steps are walked, minute-by-minute, If the person did not, there is no record
steps.utime.local Date-time minute at which steps are walked in local time
steps Amount of steps walked
decision.index Identifier for suggestions.csv given the same decision slot. Decision.index is the last notification sent, so this identifier aggregates steps walked after a notification (772 NA (= before the first notif))
slot Decision time slot for this following decision index corresponding to sugg.select.slot
study.day.nogap Day in study for decision.index, starting from 0, excluding travel days
decision.index.following.dec Identifier for suggestions.csv given the following decision slot. Decision.index.following.dec is the next notification sent, so this identifier aggregates steps walked before a notification (2421 NA (= after the last notif))
slot.following.dec Decision time slot for the following decision index corresponding to sugg.select.slot
study.day.nogap.following.dec Day in study for decision.index.following.dec, starting from 0, excluding travel days