,
Jennifer Schrack [aut] ,
Jacek Urbanek [aut] | Title: | Processing and Physical Activity Summaries of Minute Level Activity Data |
|---|---|
| Description: | Provides functions to process minute level actigraphy-measured activity counts data and extract commonly used physical activity volume and fragmentation metrics. |
| Authors: | Marta Karas [aut, cre] ,
Jennifer Schrack [aut] ,
Jacek Urbanek [aut] |
| Maintainer: | Marta Karas <[email protected]> |
| License: | GPL-3 |
| Version: | 1.1.6 |
| Built: | 2025-02-08 04:53:26 UTC |
| Source: | https://github.com/martakarass/arctools |
arctools provides 'activity_stats()' function to process minute level actigraphy-measured activity counts data and extract commonly used physical activity volume and fragmentation metrics.
To learn more about arctools, start with the vignettes: 'browseVignettes(package = "arctools")'
Maintainer: Marta Karas [email protected] (ORCID)
Authors:
Process minute level actigraphy-measured activity counts data and extract commonly used physical activity volume and fragmentation metrics.
activity_stats( acc, acc_ts, impute_missing = TRUE, sedentary_thresh = 1853, nonwear_0s_minimum_window = 90, validday_nonwear_maximum_window = 144, subset_minutes = NULL, exclude_minutes = NULL, subset_weekdays = NULL, in_bed_time = NULL, out_bed_time = NULL, adjust_out_colnames = TRUE )activity_stats( acc, acc_ts, impute_missing = TRUE, sedentary_thresh = 1853, nonwear_0s_minimum_window = 90, validday_nonwear_maximum_window = 144, subset_minutes = NULL, exclude_minutes = NULL, subset_weekdays = NULL, in_bed_time = NULL, out_bed_time = NULL, adjust_out_colnames = TRUE )
acc |
A numeric vector. A minute-level activity counts data vector. |
acc_ts |
A POSIXct vector. A minute-level time of |
impute_missing |
A logical scalar. Whether or not to perform missing
data imputation (see Details). Default is |
sedentary_thresh |
A numeric scalar. If an activity count value falls
below it then a corresponding minute is characterized as sedentary; otherwise,
a corresponding minute is characterized as active. Default is |
nonwear_0s_minimum_window |
A numeric scalar. A minimum number of consecutive minutes with 0 activity count to be considered non-wear. |
validday_nonwear_maximum_window |
In integer scalar. Maximum number of minutes of non-wear/not
collecting data so as the day is still considered valid. Default is |
subset_minutes |
Integer vector. Contains subset of a day's minutes
within which activity summaries are to be computed. May take values from
|
exclude_minutes |
Integer vector. Contains subset of a day's minutes
to be excluded from activity summaries computation.
May take values from
|
subset_weekdays |
Integer vector. Specfies days of a week within which
activity summaries are to be computed. Takes values between 1 (Sunday) to
7 (Saturday). Default is |
in_bed_time |
A POSIXct vector. An estimated in-bed time start.
Together with a corresponding entry from |
out_bed_time |
A POSIXct vector. An estimated in-bed time end.
Together with a corresponding entry from |
adjust_out_colnames |
A logical scalar. Whether or not to
add an informative suffix to column names in the output data frame.
This may happen in case
any of the arguments:
|
Physical activity statistics are aggregated from "valid" days, i.e. days
with no more than 10
wear/non-wear detection algorithm closely
following that of Choi et al. (2011).
See arctools::get_wear_flag() for details.
Data imputation is recommended for valid days for non-wear time periods and
is a default setting (see impute_missing arg).
Count values are imputed
from an "average day profile" – a minute-specific activity counts average
computed across valid days within wear time.
A data frame with physical activity summaries of minute level activity data. See README or vignette for summaries description.
Varma, V. R., Dey, D., Leroux, A., Di, J., Urbanek, J., Xiao, L., Zipunnikov, V. (2018). Total volume of physical activity: TAC, TLAC or TAC(lambda). Preventive medicine, 106, 233–235. https://doi.org/10.1016/j.ypmed.2017.10.028
Di, J., Leroux, A., Urbanek, J., Varadhan, R., Spira, A., Schrack, J., Zipunnikov, V. Patterns of sedentary and active time accumulation are associated with mortality in US adults: The NHANES study. https://doi.org/10.1101/182337
Choi, L., Liu, Z., Matthews, C. E., & Buchowski, M. S. (2011). Validation of accelerometer wear and nonwear time classification algorithm. Medicine and Science in Sports and Exercise. https://doi.org/10.1249/MSS.0b013e3181ed61a3
Koster, A., Shiroma, E. J., Caserotti, P., Matthews, C. E., Chen, K. Y., Glynn, N. W., & Harris, T. B. (2016). Comparison of Sedentary Estimates between activPAL and Hip- and Wrist-Worn ActiGraph. Medicine and science in sports and exercise, 48(8), 1514–1522. https://doi.org/10.1249/MSS.0000000000000924
fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Example 1 ## Summarize PA activity_stats(acc, acc_ts) ## Example 2 ## Summarize PA within minutes range corresponding to 12:00 AM - 6:00 AM subset_12am_6am <- 1 : (6 * 1440/24) activity_stats(acc, acc_ts, subset_minutes = subset_12am_6am) ## Example 3 ## Summarize PA without (i.e., excluding) minutes range corresponding to 11:00 PM - 5:00 AM. subset_11pm_5am <- c( (23 * 1440/24 + 1) : 1440, ## 11:00 PM - midnight 1 : (5 * 1440/24) ## midnight - 5:00 AM ) activity_stats(acc, acc_ts, exclude_minutes = subset_11pm_5am)fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Example 1 ## Summarize PA activity_stats(acc, acc_ts) ## Example 2 ## Summarize PA within minutes range corresponding to 12:00 AM - 6:00 AM subset_12am_6am <- 1 : (6 * 1440/24) activity_stats(acc, acc_ts, subset_minutes = subset_12am_6am) ## Example 3 ## Summarize PA without (i.e., excluding) minutes range corresponding to 11:00 PM - 5:00 AM. subset_11pm_5am <- c( (23 * 1440/24 + 1) : 1440, ## 11:00 PM - midnight 1 : (5 * 1440/24) ## midnight - 5:00 AM ) activity_stats(acc, acc_ts, exclude_minutes = subset_11pm_5am)
Names of exemplary accelerometry data files attached to the package.
These data can be accessed via
system.file("extdata", "<data_file_name>.csv", package = "arctools").
extdata_fnamesextdata_fnames
A character vector.
Read Actigraph device serial number from ActiLife accelerometry data file.
get_actigraph_SN(fpath_full)get_actigraph_SN(fpath_full)
fpath_full |
A string scalar. An absolute path to ActiLife accelerometry data file. |
String scalar. Actigraph device serial number.
fpath_full_i <- system.file("extdata", extdata_fnames[1], package = "arctools") get_actigraph_SN(fpath_full_i)fpath_full_i <- system.file("extdata", extdata_fnames[1], package = "arctools") get_actigraph_SN(fpath_full_i)
Compute valid/non-valid day flag (1/0) for each minute of activity
counts data.
get_valid_day_flag(wear_flag, validday_nonwear_maximum_window = 144)get_valid_day_flag(wear_flag, validday_nonwear_maximum_window = 144)
wear_flag |
An integer vector. The vector has value Vector |
validday_nonwear_maximum_window |
In integer scalar. Maxmimum number of minutes of non-wear/not
collecting data so as the day is still considered valid. Default is |
All minute-level observations
from one day are assigned the same value of valid day flag.
The flag is 1 if a day is
determined to be valid, and 0 otherwise.
A day is determined to be valid if it has no more than
validday_nonwear_maximum_window minutes of missing data.
Data may be missing due to
identified sensor nonwear or because activity data collection has not started
yet/has finished already in a particular day.
An integer vector. It has value 1 if a minute
belongs to a valid day, and 0 otherwise.
## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag)## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag)
Compute wear/non-wear flag (1/0) for each minute of activity
counts data.
get_wear_flag(acc, nonwear_0s_minimum_window = 90)get_wear_flag(acc, nonwear_0s_minimum_window = 90)
acc |
A numeric vector. A minute-level activity counts data vector. |
nonwear_0s_minimum_window |
A numeric scalar. A minimum number of consecutive minutes with 0 activity count to be considered non-wear. |
Method implements wear/non-wear detection algorithm closely following that of Choi et al. (2011).
The wear/non-wear flag is determined based on activity counts data.
A minute is classified as non-wear if it belongs to any
nonwear_0s_minimum_window minutes-long interval of consecutive values 0 in
activity counts data vector;
here, "any interval" implies that a particular minute may be located
at any location (beginning, middle, end) of interval of consecutive values 0
to be classified as a non-wear. Otherwise, a particular minute is classified
as wear.
Similarly to recommendations in Discussion in Choi et al. (2011), the method assumes a threshold value of 0 for nonzero counts allowed during a nonwear time interval (I.e., no activity count equal >= 1 is allowed). The method also assumes 90 minutes as a default for minimum time of consecutive zero counts for a window to be flagged nonwear. Differently from recommendations in Discussion in Choi et al. (2011), it does not consider any "artifactual movement" interval of nonzero counts during a nonwear time interval.
An integer vector. It has value 1 for a wear
and 0 for non-wear flagged minute. It has the same vector length as
acc vector. If there is an NA entry in acc vector,
then the returned vector will have a corresponding entry set to NA too.
Choi, L., Liu, Z., Matthews, C. E., & Buchowski, M. S. (2011). Validation of accelerometer wear and nonwear time classification algorithm. Medicine and Science in Sports and Exercise. https://doi.org/10.1249/MSS.0b013e3181ed61a3
## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc)## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc)
Impute missing data in minute-level activity counts data vector based on "average day profile".
impute_missing_data( acc, wear_flag, valid_day_flag, imputeFromValidDaysOnly = TRUE )impute_missing_data( acc, wear_flag, valid_day_flag, imputeFromValidDaysOnly = TRUE )
acc |
A numeric vector. A minute-level activity counts data vector.
It is assumed
to be in midnight-to-midnight format, meaning its vector length
is a multiple of number of minutes in a full day
(1440; see |
wear_flag |
An integer vector.
Wear/non-wear flag ( |
valid_day_flag |
An integer vector.
Valid/non-valid day flag ( |
imputeFromValidDaysOnly |
A logical scalar. Whether or not data from valid days only should be used for computing "average day profile" used for imputation. |
An "average day profile" is computed as average across minutes identified
as wear and from valid days
(see param. imputeFromValidDaysOnly). Activity counts data are imputed
from "average day profile" for minutes identified as non-wear in days
identified as valid, except for minutes before/after data
collection start/end which remain NA.
Theoretically, it is possible that all valid days of data collection have non-wear flag for the some minute(s) (i.e., somebody is always taking off the watch for the same few minutes during a day) so there is no data to use to compute imputation values from. If it happens, then method uses 0 as imputation value(s).
A numeric vector. A minute-level activity counts data vector with data imputed for minutes identified as non-wear in days identified as valid
## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag) ## Impute missing data in acc data vector acc_imputed <- impute_missing_data(acc, wear_flag, valid_day_flag) ## Compare mean acc before/after imputation c(mean(acc, na.rm = TRUE), mean(acc_imputed, na.rm = TRUE))## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag) ## Impute missing data in acc data vector acc_imputed <- impute_missing_data(acc, wear_flag, valid_day_flag) ## Compare mean acc before/after imputation c(mean(acc, na.rm = TRUE), mean(acc_imputed, na.rm = TRUE))
Expand activity data vector such that its length is a multiple of number of minutes in a full day (1440).
midnight_to_midnight(acc, acc_ts)midnight_to_midnight(acc, acc_ts)
acc |
A numeric vector. A minute-level activity counts data vector. |
acc_ts |
A POSIXct vector. Time of activity data collection, corresponding to
|
In the returned vector,
1st observation
corresponds to minute of 00:00-00:01 on the first day of data collection,
and last observation
corresponds to minute of 23:59-00:00 on the last day of data collection.
Entries corresponding to no data in original activity data vector
are filled with NA.
A numeric vector. A minute-level activity counts data vector in midnight-to-midnight format.
## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Observe we have an integer number of days length(acc) / 1440## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Observe we have an integer number of days length(acc) / 1440
Compute physical activity summaries of minute level activity data.
summarize_PA( acc, acc_ts, wear_flag, valid_day_flag, sedentary_thresh = 1853, subset_minutes = NULL, exclude_minutes = NULL, subset_weekdays = NULL, in_bed_time = NULL, out_bed_time = NULL, adjust_out_colnames = TRUE )summarize_PA( acc, acc_ts, wear_flag, valid_day_flag, sedentary_thresh = 1853, subset_minutes = NULL, exclude_minutes = NULL, subset_weekdays = NULL, in_bed_time = NULL, out_bed_time = NULL, adjust_out_colnames = TRUE )
acc |
A numeric vector. A minute-level activity counts data vector.
It is assumed
to be in midnight-to-midnight format, meaning its vector length
is a multiple of number of minutes in a full day
(1440). See |
acc_ts |
A POSIXct vector. Time of activity data collection, corresponding to
|
wear_flag |
An integer vector. It has value |
valid_day_flag |
An integer vector. It has value |
sedentary_thresh |
A numeric scalar. If an activity count value falls
below it then a corresponding minute is characterized as sedentary; otherwise,
a corresponding minute is characterized as active. Default is |
subset_minutes |
Integer vector. Contains subset of a day's minutes
within which activity summaries are to be computed. May take values from
|
exclude_minutes |
Integer vector. Contains subset of a day's minutes
to be excluded from activity summaries computation.
May take values from
|
subset_weekdays |
Integer vector. Specfies days of a week within which
activity summaries are to be computed. Takes values between 1 (Sunday) to
7 (Saturday). Default is |
in_bed_time |
A POSIXct vector. An estimated in-bed time start.
Together with a corresponding entry from |
out_bed_time |
A POSIXct vector. An estimated in-bed time end.
Together with a corresponding entry from |
adjust_out_colnames |
A logical scalar. Whether or not to
add an informative suffix to column names in the output data frame.
This may happen in case
any of the arguments:
|
A data frame with physical activity summaries of minute level activity data. See README or vignette for summaries description.
## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag) ## Impute missing data in acc data vector acc_imputed <- impute_missing_data(acc, wear_flag, valid_day_flag) ## Example 1 ## Summarize PA summarize_PA(acc, acc_ts, wear_flag, valid_day_flag) ## Example 2 ## Summarize PA within minutes range corresponding to 12:00 AM - 6:00 AM subset_12am_6am <- 1 : (6 * 1440/24) summarize_PA(acc, acc_ts, wear_flag, valid_day_flag, subset_minutes = subset_12am_6am) ## Example 3 ## Summarize PA without (i.e., excluding) minutes range corresponding to 11:00 PM - 5:00 AM. subset_11pm_5am <- c( (23 * 1440/24 + 1) : 1440, ## 11:00 PM - midnight 1 : (5 * 1440/24) ## midnight - 5:00 AM ) summarize_PA(acc, acc_ts, wear_flag, valid_day_flag, exclude_minutes = subset_11pm_5am)## Read exemplary data fpath_i <- system.file("extdata", extdata_fnames[1], package = "arctools") dat_i <- as.data.frame(data.table::fread(fpath_i)) acc <- dat_i$vectormagnitude acc_ts <- lubridate::ymd_hms(dat_i$timestamp) ## Get acc data vector in "midnight_to_midnight" format acc <- midnight_to_midnight(acc, acc_ts) ## Get wear/non-wear flag wear_flag <- get_wear_flag(acc) ## Get valid/non-valid day flag valid_day_flag <- get_valid_day_flag(wear_flag) ## Impute missing data in acc data vector acc_imputed <- impute_missing_data(acc, wear_flag, valid_day_flag) ## Example 1 ## Summarize PA summarize_PA(acc, acc_ts, wear_flag, valid_day_flag) ## Example 2 ## Summarize PA within minutes range corresponding to 12:00 AM - 6:00 AM subset_12am_6am <- 1 : (6 * 1440/24) summarize_PA(acc, acc_ts, wear_flag, valid_day_flag, subset_minutes = subset_12am_6am) ## Example 3 ## Summarize PA without (i.e., excluding) minutes range corresponding to 11:00 PM - 5:00 AM. subset_11pm_5am <- c( (23 * 1440/24 + 1) : 1440, ## 11:00 PM - midnight 1 : (5 * 1440/24) ## midnight - 5:00 AM ) summarize_PA(acc, acc_ts, wear_flag, valid_day_flag, exclude_minutes = subset_11pm_5am)