bweiher/pelotonr

R package that interfaces with the Peloton API

pelotonr's Issues

shiny dashboard

it would be cool to create a nice shiny dashboard that could:

• update gathered data (maybe add to a local data store or DB ; incrementally find new data it needs)

• create some plots based off data
  show user progress

• activity patterns, etc

issue with joins in get_all_workouts()

library(pelotonR)
library(tidyverse)


Sys.setenv("PELOTON_LOGIN" = "username")
Sys.setenv("PELOTON_PASSWORD" = "pw")

peloton_auth()

# ----------------------- get data about yourself
me <- get_my_info() # peloton_api("api/me")
user_id <- me$id

# ---------------- GET

pel_workouts <- get_all_workouts(user_id, num_workouts = me$total_workouts, joins = "ride,ride.instructor")

Error: Problem with `mutate()` input `live_stream_id`.
x character string is not in a standard unambiguous format
i Input `live_stream_id` is `(function (x, ...) ...`.

Getting this error when using joins = "ride,ride.instructor"

Error when retrieving workouts

I'm getting an error when I try to retrieve my workouts:

library(pelotonR)
peloton_auth()
#> [1] "Logged in"

# get data about yourself
me <- get_my_info() # peloton_api("api/me")
user_id <- me$id

# get a list of your workouts
workouts <- get_all_workouts(user_id) # peloton_api("api/user_id/workouts")
#> Error: Can't combine `..1$v2_total_video_watch_time_seconds` <integer> and `..18$v2_total_video_watch_time_seconds` <character>.

Allow for paging when querying workouts

It appears there is a limit of 100 responses when querying workouts, which means you have to page through in batches of 100 to get all workouts (if you have more than 100). Right now page=0 is hardcoded into the function, but it seems like there should be a way to make it so you can specify the page. I am currently using this, where the only thing I'm modiying is the page_num in the parameters and API call.

get_all_workouts_page <- function(userid = Sys.getenv("PELOTON_USERID"),
                             num_workouts = 20,
                             joins = "",
                             page_num = 0, 
                             dictionary = list(
                                 "numeric" = c(
                                     "v2_total_video_buffering_seconds",
                                     "v2_total_video_watch_time_seconds"
                                 )
                             ),
                             parse_dates = TRUE,
                             ...) {
    if (userid == "")
        stop("Provide a userid or set an environmental variable `PELOTON_USERID`",
             call. = FALSE)
    if (length(joins) > 1 ||
        !is.character(joins))
        stop("Provide joins as a length one character vector", call. = FALSE)
    
    # see if joins is provided, if so, append to request
    if (joins != "")
        joins <- glue::glue("joins={joins}")
    
    workouts <-
        peloton_api(
            glue::glue(
                "/api/user/{userid}/workouts?{joins}&limit={num_workouts}&page={page_num}"
            ),
            ...
        )
    n_workouts <- length(workouts$content$data)
    # v2_total_video_buffering_seconds v2_total_video_watch_time_seconds
    if (n_workouts > 0) {
        workouts <-
            purrr::map_df(
                1:n_workouts,
                ~ parse_list_to_df(
                    workouts$content$data[[.]],
                    dictionary = dictionary,
                    parse_dates = parse_dates
                )
            )
        
        # IF JOIN PARAM is specified, get data out for ride list and add it to that row
        if (joins != "") {
            rides <- purrr::map_df(1:n_workouts, function(x) {
                parse_list_to_df(
                    stats::setNames(workouts$ride[[x]], paste0(
                        "ride_", names(workouts$ride[[x]])
                    )),
                    dictionary = dictionary,
                    parse_dates = parse_dates,
                    ...
                )
            })
            
            dplyr::left_join(
                dplyr::mutate(workouts, rn = dplyr::row_number()),
                dplyr::mutate(rides, rn = dplyr::row_number()),
                by  = "rn"
            ) %>%
                dplyr::select(-.data$rn)
        } else {
            workouts
        }
    }
}

query caching

if you run get_perfomance_graphs (maybe get_all_workouts, but would have to be smart) repeatedly you may have to fetch and process the same data many times.

its possible to cache which workouts already have fetched data, so it doesnt hit the API. older results could be stored in local files.

Thank you

I just discovered today I can download my peloton data. then through lucky googling I found this package. Was worried I'd have to do this myself... Thanks for your work here - you've made my day!