Merge pull request #155 from olivroy/upkeep2

.Rbuildignore

@@ -3,7 +3,6 @@
 ^data-raw$
 ^README\.Rmd$
 ^README-.*\.png$
-^\.travis\.yml$
 ^cran-comments\.md$
 ^revdep$
 img

.gitignore

@@ -9,3 +9,4 @@ Meta
 docs
 /doc/
 /Meta/
+docs/

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: timetk
 Type: Package
 Title: A Tool Kit for Working with Time Series
-Version: 2.8.4
+Version: 2.8.4.9000
 Authors@R: c(
     person("Matt", "Dancho", email = "[email protected]", role = c("aut", "cre")),
     person("Davis", "Vaughan", email = "[email protected]", role = c("aut"))
@@ -43,15 +43,13 @@ Imports:
     forecast,
     tsfeatures,
     hms,
-    generics,
-    tidymodels
+    generics
 Suggests:
     modeltime,
     glmnet,
     workflows,
     parsnip,
-    tune,
-    yardstick,
+    tune (>= 0.1.2),
     knitr,
     rmarkdown,
     broom,

NEWS.md

@@ -1,3 +1,8 @@
+# timetk (development version)
+
+- Removed dependency on tidymodels. (#154, @olivroy).
+
+
 # timetk 2.8.4
 
 Update forecasting vignette: Use `glmnet` for time series forecasting. 

R/augment-tk_augment_differences.R

@@ -43,7 +43,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' m4_monthly %>%
 #'     group_by(id) %>%

R/augment-tk_augment_fourier.R

@@ -41,7 +41,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' m4_monthly %>%
 #'     group_by(id) %>%
@@ -65,7 +64,7 @@ tk_augment_fourier <- function(.data,
     # if (rlang::is_missing(.K)) stop(call. = FALSE, "tk_augment_fourier(.K) is missing.")
     if (!any(.names == "auto")) {
         if (length(.names) != length(.periods) * 2) {
-            rlang::abort(".names must be a vector of length ", length(.periods) * 2)
+            rlang::abort(c(".names must be a vector of length ", length(.periods) * 2))
         }
     }
     vals <- .data %>% dplyr::pull(!! rlang::enquo(.date_var))

R/augment-tk_augment_holiday_signature.R

@@ -55,7 +55,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' dates_in_2017_tbl <- tibble(index = tk_make_timeseries("2017-01-01", "2017-12-31", by = "day"))
 #'
@@ -137,11 +136,10 @@ tk_augment_holiday_signature.data.frame <- function(.data,
     # .data <- .data %>% dplyr::arrange(!! sym(date_var))
 
     # Bind Time Series Signature
-    ret_1 <- .data %>% dplyr::ungroup()
+    ret_1 <- dplyr::ungroup(.data)
 
-    ret_2 <- .data %>%
-        dplyr::ungroup() %>%
-        dplyr::pull(date_var) %>%
+    ret_2 <- ret_1 %>%
+        dplyr::pull(dplyr::all_of(date_var)) %>%
         tk_get_holiday_signature(
             holiday_pattern = .holiday_pattern,
             locale_set      = .locale_set,

R/augment-tk_augment_lags.R

@@ -46,7 +46,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' # Lags
 #' m4_monthly %>%

R/augment-tk_augment_slidify.R

@@ -42,8 +42,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(stringr)
-#' library(timetk)
 #'
 #' # Single Column | Multiple Rolling Windows
 #' FANG %>%
@@ -55,7 +53,7 @@
 #'         .period  = c(10, 30, 60, 90),
 #'         .f       = mean,
 #'         .partial = TRUE,
-#'         .names   = str_c("MA_", c(10, 30, 60, 90))
+#'         .names   = stringr::str_c("MA_", c(10, 30, 60, 90))
 #'     ) %>%
 #'     ungroup()
 #'
@@ -184,7 +182,7 @@ tk_augment_slidify.grouped_df <- function(.data,
                 .names     = .names
             )
         )) %>%
-        dplyr::select(-data) %>%
+        dplyr::select(-"data") %>%
         tidyr::unnest(cols = nested.col) %>%
         dplyr::group_by_at(.vars = group_names)
 }

R/augment-tk_augment_timeseries.R

@@ -32,7 +32,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' m4_daily %>%
 #'     group_by(id) %>%
@@ -109,7 +108,7 @@ tk_augment_timeseries_signature.grouped_df <- function(.data, .date_var = NULL)
                     .date_var   = NULL
                 )
             )) %>%
-            dplyr::select(-data) %>%
+            dplyr::select(-"data") %>%
             tidyr::unnest(cols = nested.col) %>%
             dplyr::group_by_at(.vars = group_names)
     } else {
@@ -122,7 +121,7 @@ tk_augment_timeseries_signature.grouped_df <- function(.data, .date_var = NULL)
                     .date_var   = !! date_var_expr
                 )
             )) %>%
-            dplyr::select(-data) %>%
+            dplyr::select(-"data") %>%
             tidyr::unnest(cols = nested.col) %>%
             dplyr::group_by_at(.vars = group_names)
     }

R/coersion-tk_tbl.R

@@ -31,9 +31,7 @@
 #' @seealso [tk_xts()], [tk_zoo()], [tk_zooreg()], [tk_ts()]
 #'
 #' @examples
-#' library(tibble)
 #' library(dplyr)
-#' library(timetk)
 #'
 #' data_tbl <- tibble(
 #'     date = seq.Date(from = as.Date("2010-01-01"), by = 1, length.out = 5),
@@ -85,7 +83,6 @@
 #'
 #'
 #'
-#'
 #' @export
 tk_tbl <- function(data, preserve_index = TRUE, rename_index = "index", timetk_idx = FALSE, silent = FALSE, ...) {
     UseMethod("tk_tbl", data)

R/coersion-tk_ts.R

@@ -31,9 +31,7 @@
 #' @seealso [tk_index()], [tk_tbl()], [tk_xts()], [tk_zoo()], [tk_zooreg()]
 #'
 #' @examples
-#' library(tibble)
 #' library(dplyr)
-#' library(timetk)
 #'
 #' ### tibble to ts: Comparison between tk_ts() and stats::ts()
 #' data_tbl <- tibble::tibble(

R/coersion-tk_xts.R

@@ -38,9 +38,7 @@
 #' @seealso [tk_tbl()], [tk_zoo()], [tk_zooreg()], [tk_ts()]
 #'
 #' @examples
-#' library(tibble)
 #' library(dplyr)
-#' library(timetk)
 #'
 #' ### tibble to xts: Comparison between tk_xts() and xts::xts()
 #' data_tbl <- tibble::tibble(
@@ -69,7 +67,6 @@
 #' select   <- "y"
 #' tk_xts_(data_tbl, select = select, date_var = date_var)
 #'
-#' @name tk_xts
 #' @export
 tk_xts <- function(data, select = NULL, date_var = NULL, silent = FALSE, ...) {
     select   <- rlang::quo_name(rlang::enquo(select))

R/coersion-tk_zoo.R

@@ -40,12 +40,10 @@
 #' @seealso [tk_tbl()], [tk_xts()], [tk_zooreg()], [tk_ts()]
 #'
 #' @examples
-#' library(tibble)
 #' library(dplyr)
-#' library(timetk)
 #'
 #' ### tibble to zoo: Comparison between tk_zoo() and zoo::zoo()
-#' data_tbl <- tibble::tibble(
+#' data_tbl <- dplyr::tibble(
 #'     date = seq.Date(as.Date("2016-01-01"), by = 1, length.out = 5),
 #'     x    = rep("chr values", 5),
 #'     y    = cumsum(1:5),

R/diagnostics-tk_acf_diagnostics.R

@@ -61,7 +61,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' # ACF, PACF, & CCF in 1 Data Frame
 #' # - Get ACF & PACF for target (adjusted)
@@ -139,7 +138,7 @@ tk_acf_diagnostics.data.frame <- function(.data, .date_var, .value, .ccf_vars =
 
     # Calcs
     .lags   <- sort(.lags)
-    x       <- .data %>% dplyr::pull(.value_mod)
+    x       <- .data %>% dplyr::pull(".value_mod")
     lag_max <- max(.lags)
     lag_min <- min(.lags)
 
@@ -239,7 +238,7 @@ tk_acf_diagnostics.grouped_df <- function(.data, .date_var, .value, .ccf_vars =
                 .lags      = .lags
             )
         )) %>%
-        dplyr::select(-data) %>%
+        dplyr::select(-"data") %>%
         tidyr::unnest(cols = nested.col) %>%
         dplyr::group_by_at(.vars = group_names)
 

R/diagnostics-tk_anomaly_diagnostics.R

@@ -78,7 +78,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' walmart_sales_weekly %>%
 #'     filter(id %in% c("1_1", "1_3")) %>%
@@ -178,7 +177,7 @@ tk_anomaly_diagnostics.grouped_df <- function(.data, .date_var, .value,
                 .message       = .message
             )
         )) %>%
-        dplyr::select(-data) %>%
+        dplyr::select(-"data") %>%
         tidyr::unnest(cols = nested.col) %>%
         dplyr::group_by_at(.vars = group_names)
 
@@ -294,15 +293,15 @@ iqr_vec <- function(x, alpha = 0.05, max_anoms = 0.2, verbose = FALSE) {
     # Return results
     if (verbose) {
         outlier_list <- list(
-            outlier = vals_tbl %>% dplyr::arrange(index) %>% dplyr::pull(outlier_reported),
-            outlier_idx = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull(index),
-            outlier_vals = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull(value),
-            outlier_direction = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull(direction),
+            outlier = vals_tbl %>% dplyr::arrange(index) %>% dplyr::pull("outlier_reported"),
+            outlier_idx = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull("index"),
+            outlier_vals = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull("value"),
+            outlier_direction = vals_tbl %>% dplyr::filter(outlier_reported == "Yes") %>% dplyr::pull("direction"),
             critical_limits = limits_vec,
             outlier_report = vals_tbl_filtered
         )
         return(outlier_list)
     } else {
-        return(vals_tbl %>% dplyr::arrange(index) %>% dplyr::pull(outlier_reported))
+        return(vals_tbl %>% dplyr::arrange(index) %>% dplyr::pull("outlier_reported"))
     }
 }

R/diagnostics-tk_seasonal_diagnostics.R

@@ -48,7 +48,6 @@
 #' @examples
 #' \donttest{
 #' library(dplyr)
-#' library(timetk)
 #'
 #' # ---- GROUPED EXAMPLES ----
 #'
@@ -170,7 +169,7 @@ tk_seasonal_diagnostics.grouped_df <- function(.data, .date_var, .value, .featur
                 .feature_set  = .feature_set
             )
         )) %>%
-        dplyr::select(-data) %>%
+        dplyr::select(-"data") %>%
         tidyr::unnest(cols = nested.col) %>%
         dplyr::group_by_at(.vars = group_names)
 
@@ -194,7 +193,7 @@ get_seasonal_auto_features <- function(.index) {
         dplyr::mutate(check = value %>% dplyr::between(max_min_list$min_period$value, max_min_list$max_period$value)) %>%
         dplyr::filter(check) %>%
         dplyr::left_join(time_series_signature_lookup_tbl(), by = "key") %>%
-        dplyr::pull(feature)
+        dplyr::pull("feature")
 
     return(features_to_get)
 }
@@ -219,7 +218,7 @@ get_max_min_list <- function(time_series_summary_tbl) {
         tidyr::gather() %>%
         dplyr::mutate(check = 2 * value < start_to_end) %>%
         dplyr::filter(check) %>%
-        dplyr::slice(dplyr::n())
+        dplyr::slice_tail(n = 1)
 
     # Max and min
     max_min_list <- list(min_period = min_period, max_period = max_period)

R/diagnostics-tk_stl_diagnostics.R

@@ -47,7 +47,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #'
 #' # ---- GROUPS & TRANSFORMATION ----
@@ -112,7 +111,7 @@ tk_stl_diagnostics.data.frame <- function(.data, .date_var, .value,
 
     # STL Calculation
     stl_obj <- data_formatted %>%
-        dplyr::pull(.value_mod) %>%
+        dplyr::pull(".value_mod") %>%
         stats::ts(frequency = freq) %>%
         stats::stl(s.window = "periodic", t.window = trnd, robust = TRUE)
 

R/diagnostics-tk_summary_diagnostics.R

@@ -19,7 +19,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' # ---- NON-GROUPED EXAMPLES ----
 #'

R/diagnostiscs-tsfeatures.R

@@ -71,7 +71,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' walmart_sales_weekly %>%
 #'     group_by(id) %>%

R/dplyr-add_time.R

@@ -36,7 +36,6 @@
 #'
 #'
 #' @examples
-#' library(timetk)
 #'
 #'
 #' # ---- LOCATING A DATE N-PERIODS IN FUTURE / PAST ----
@@ -67,7 +66,7 @@
 #'
 #'
 #' @name time_arithmetic
-#' @export
+NULL
 
 # ADD ----
 

R/dplyr-between_time.R

@@ -65,7 +65,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' index_daily <- tk_make_timeseries("2016-01-01", "2017-01-01", by = "day")
 #' index_min   <- tk_make_timeseries("2016-01-01", "2017-01-01", by = "min")

R/dplyr-condense_period.R

@@ -54,7 +54,6 @@
 #'
 #' @examples
 #' # Libraries
-#' library(timetk)
 #' library(dplyr)
 #'
 #' # First value in each month

R/dplyr-filter_by_time.R

@@ -65,7 +65,6 @@
 #'
 #' @examples
 #' library(dplyr)
-#' library(timetk)
 #'
 #' # Filter values in January 1st through end of February, 2013
 #' FANG %>%

R/dplyr-filter_period.R

@@ -58,7 +58,6 @@
 #'
 #' @examples
 #' # Libraries
-#' library(timetk)
 #' library(dplyr)
 #'
 #' # Max value in each month