twang2218 / pmap / #93

pending completion

Build # #93

Build Type

push

jenkins

Committed by

web-flow

Commit Message

<a href="https://github.com/twang2218/pmap/commit/<a class=hub.com/twang2218/pmap/commit/19f030a29116d7213aca66d9ca167a8682708987">19f030a29<a href="https://github.com/twang2218/pmap/commit/19f030a29116d7213aca66d9ca167a8682708987">">Merge </a><a class="double-link" href="https://github.com/twang2218/pmap/commit/<a class="double-link" href="https://github.com/twang2218/pmap/commit/2f405da75d783698b6cb42985b748a7023953d21">2f405da75</a>">2f405da75</a><a href="https://github.com/twang2218/pmap/commit/19f030a29116d7213aca66d9ca167a8682708987"> into 80b26a7f2">80b26a7f2</a>

Run Details

80 of 80 new or added lines in 4 files covered. (100.0%)

414 of 434 relevant lines covered (95.39%)

26.86 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

90.48

/R/create_pmap.R

#' @title Create the process map from event log directly
#' @param eventlog Event log
#' @param distinct_case Whether should count distinct case only. Default is `FALSE`.
#' @param distinct_repeated_activities Whether should distinct repeat activities. Default is `FALSE`, which means the repeated activity will be treated as same node. If it's `TRUE`, the name of the activity will be attached with the sequence number of occurance of the activity.
#' @param target_categories A vector contains the target activity categories
#' @param edge_label Specify which attribute is used for the edge label.
#' @param edge_width Specify which attribute is used for the edge width.
#' @description Create the process map by analyzing the given `eventlog` and extract the nodes by `generate_nodes()` and edges by `generate_edges()`.
#' @details
#' ```R
#' > eventlog <- data.frame(
#'     timestamp = c(
#'       as.POSIXct("2017-10-01"),
#'       as.POSIXct("2017-10-02"),
#'       as.POSIXct("2017-10-03"),
#'       as.POSIXct("2017-10-04"),
#'       as.POSIXct("2017-10-05"),
#'       as.POSIXct("2017-10-06"),
#'       as.POSIXct("2017-10-07"),
#'       as.POSIXct("2017-10-08"),
#'       as.POSIXct("2017-10-09"),
#'       as.POSIXct("2017-10-10")
#'     ),
#'     case_id = c("c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1"),
#'     activity =  c("a",  "b",  "d",  "a",  "c",  "a",  "b",  "c",  "a",  "d"),
#'     category =  c("campaign", "campaign", "sale", "campaign", "sale", "campaign", "campaign", "sale", "campaign", "sale"),
#'     stringsAsFactors = FALSE
#'   )
#' > eventlog
#'     timestamp case_id activity category
#' 1  2017-10-01          c1          a   campaign
#' 2  2017-10-02          c1          b   campaign
#' 3  2017-10-03          c1          d       sale
#' 4  2017-10-04          c1          a   campaign
#' 5  2017-10-05          c1          c       sale
#' 6  2017-10-06          c1          a   campaign
#' 7  2017-10-07          c1          b   campaign
#' 8  2017-10-08          c1          c       sale
#' 9  2017-10-09          c1          a   campaign
#' 10 2017-10-10          c1          d       sale
#' > p <- create_pmap(eventlog, target_categories = c("sale"))
#' > render_pmap(p)
#' ```
#' \if{html}{\figure{example.create_pmap.simple.svg}{options: alt="Figure: example.create_pmap.simple.svg"}}
#'
#' Or for more complex event log:
#'
#' ```R
#' > eventlog <- generate_eventlog(
#'     size_of_eventlog = 10000,
#'     number_of_cases = 2000,
#'     categories = c("campaign", "sale"),
#'     categories_size = c(8, 2))
#' > head(eventlog)
#'             timestamp   case_id         activity category
#' 1 2017-01-01 02:40:20 Case 1204 Activity 7 (campaign)   campaign
#' 2 2017-01-01 03:10:31 Case 1554 Activity 5 (campaign)   campaign
#' 3 2017-01-01 04:01:51  Case 546 Activity 4 (campaign)   campaign
#' 4 2017-01-01 05:04:09 Case 1119     Activity 9 (sale)       sale
#' 5 2017-01-01 06:43:11 Case 1368 Activity 2 (campaign)   campaign
#' 6 2017-01-01 07:43:06  Case 986 Activity 8 (campaign)   campaign
#' > str(eventlog)
#' 'data.frame':   10000 obs. of  4 variables:
#'  $ timestamp  : POSIXct, format: "2017-01-01 02:40:20" "2017-01-01 03:10:31" ...
#'  $ case_id: chr  "Case 1204" "Case 1554" "Case 546" "Case 1119" ...
#'  $ activity : chr  "Activity 7 (campaign)" "Activity 5 (campaign)" "Activity 4 (campaign)" "Activity 9 (sale)" ...
#'  $ category : chr  "campaign" "campaign" "campaign" "sale" ...
#' > p <- create_pmap(eventlog, target_categories = c("sale"))
#' > render_pmap(p)
#' ```
#' \if{html}{\figure{example.create_pmap.complex.svg}{options: width="100\%" alt="Figure: example.create_pmap.complex.svg"}}
#'
#' @seealso [prune_edges]
#' @seealso [create_pmap_graph]
#' @importFrom dplyr        %>%
#' @importFrom dplyr        group_by
#' @importFrom dplyr        arrange
#' @importFrom dplyr        rename
#' @importFrom dplyr        mutate
#' @importFrom dplyr        ungroup
#' @importFrom stringr      str_trim
#' @export
create_pmap <- function(
  eventlog,
  distinct_case = FALSE,
  distinct_repeated_activities = FALSE,
  target_categories = NULL,
  edge_label = c(
    "amount",
    "mean_duration",
    "median_duration",
    "max_duration",
    "min_duration"
  ),
  edge_width = c(
    "amount",
    "mean_duration",
    "median_duration",
    "max_duration",
    "min_duration"
  )
) {
  # Make R Cmd Check happy
  old_activity <- case_id <- activity <- old_activity <- timestamp <- NULL

  # clean names
  eventlog <- eventlog %>%
    dplyr::mutate_if(is.factor, as.character) %>%
    dplyr::mutate_if(is.character, stringr::str_trim)

  if (distinct_repeated_activities) {
    eventlog <- eventlog %>%
      dplyr::group_by(case_id, activity) %>%
      dplyr::arrange(timestamp, .by_group = TRUE) %>%
      dplyr::rename(old_activity = activity) %>%
      dplyr::mutate(activity = paste0(old_activity, " (", 1:n(), ")")) %>%
      dplyr::ungroup()

    if (!"category" %in% colnames(eventlog)) {
      # if the `category` is missing,
      # then use the original `activity` as the `category`
      eventlog <- eventlog %>% dplyr::rename(category = old_activity)
    }
  }

  nodes <- generate_nodes(eventlog, distinct_case)
  edges <- generate_edges(eventlog, distinct_case, target_categories)

  if (nrow(nodes) == 0) {
    stop("Generated graph contains empty node.")
  }

  if (nrow(edges) == 0) {
    stop("Generated graph contains no edge.")
  }

  p <- create_pmap_graph(nodes, edges, target_categories, edge_label, edge_width)
  return(p)
}

1	#' @title Create the process map from event log directly
2	#' @param eventlog Event log
3	#' @param distinct_case Whether should count distinct case only. Default is `FALSE`.
4	#' @param distinct_repeated_activities Whether should distinct repeat activities. Default is `FALSE`, which means the repeated activity will be treated as same node. If it's `TRUE`, the name of the activity will be attached with the sequence number of occurance of the activity.
5	#' @param target_categories A vector contains the target activity categories
6	#' @param edge_label Specify which attribute is used for the edge label.
7	#' @param edge_width Specify which attribute is used for the edge width.
8	#' @description Create the process map by analyzing the given `eventlog` and extract the nodes by `generate_nodes()` and edges by `generate_edges()`.
9	#' @details
10	#' ```R
11	#' > eventlog <- data.frame(
12	#' timestamp = c(
13	#' as.POSIXct("2017-10-01"),
14	#' as.POSIXct("2017-10-02"),
15	#' as.POSIXct("2017-10-03"),
16	#' as.POSIXct("2017-10-04"),
17	#' as.POSIXct("2017-10-05"),
18	#' as.POSIXct("2017-10-06"),
19	#' as.POSIXct("2017-10-07"),
20	#' as.POSIXct("2017-10-08"),
21	#' as.POSIXct("2017-10-09"),
22	#' as.POSIXct("2017-10-10")
23	#' ),
24	#' case_id = c("c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1", "c1"),
25	#' activity = c("a", "b", "d", "a", "c", "a", "b", "c", "a", "d"),
26	#' category = c("campaign", "campaign", "sale", "campaign", "sale", "campaign", "campaign", "sale", "campaign", "sale"),
27	#' stringsAsFactors = FALSE
28	#' )
29	#' > eventlog
30	#' timestamp case_id activity category
31	#' 1 2017-10-01 c1 a campaign
32	#' 2 2017-10-02 c1 b campaign
33	#' 3 2017-10-03 c1 d sale
34	#' 4 2017-10-04 c1 a campaign
35	#' 5 2017-10-05 c1 c sale
36	#' 6 2017-10-06 c1 a campaign
37	#' 7 2017-10-07 c1 b campaign
38	#' 8 2017-10-08 c1 c sale
39	#' 9 2017-10-09 c1 a campaign
40	#' 10 2017-10-10 c1 d sale
41	#' > p <- create_pmap(eventlog, target_categories = c("sale"))
42	#' > render_pmap(p)
43	#' ```
44	#' \if{html}{\figure{example.create_pmap.simple.svg}{options: alt="Figure: example.create_pmap.simple.svg"}}
45	#'
46	#' Or for more complex event log:
47	#'
48	#' ```R
49	#' > eventlog <- generate_eventlog(
50	#' size_of_eventlog = 10000,
51	#' number_of_cases = 2000,
52	#' categories = c("campaign", "sale"),
53	#' categories_size = c(8, 2))
54	#' > head(eventlog)
55	#' timestamp case_id activity category
56	#' 1 2017-01-01 02:40:20 Case 1204 Activity 7 (campaign) campaign
57	#' 2 2017-01-01 03:10:31 Case 1554 Activity 5 (campaign) campaign
58	#' 3 2017-01-01 04:01:51 Case 546 Activity 4 (campaign) campaign
59	#' 4 2017-01-01 05:04:09 Case 1119 Activity 9 (sale) sale
60	#' 5 2017-01-01 06:43:11 Case 1368 Activity 2 (campaign) campaign
61	#' 6 2017-01-01 07:43:06 Case 986 Activity 8 (campaign) campaign
62	#' > str(eventlog)
63	#' 'data.frame': 10000 obs. of 4 variables:
64	#' $ timestamp : POSIXct, format: "2017-01-01 02:40:20" "2017-01-01 03:10:31" ...
65	#' $ case_id: chr "Case 1204" "Case 1554" "Case 546" "Case 1119" ...
66	#' $ activity : chr "Activity 7 (campaign)" "Activity 5 (campaign)" "Activity 4 (campaign)" "Activity 9 (sale)" ...
67	#' $ category : chr "campaign" "campaign" "campaign" "sale" ...
68	#' > p <- create_pmap(eventlog, target_categories = c("sale"))
69	#' > render_pmap(p)
70	#' ```
71	#' \if{html}{\figure{example.create_pmap.complex.svg}{options: width="100\%" alt="Figure: example.create_pmap.complex.svg"}}
72	#'
73	#' @seealso [prune_edges]
74	#' @seealso [create_pmap_graph]
75	#' @importFrom dplyr %>%
76	#' @importFrom dplyr group_by
77	#' @importFrom dplyr arrange
78	#' @importFrom dplyr rename
79	#' @importFrom dplyr mutate
80	#' @importFrom dplyr ungroup
81	#' @importFrom stringr str_trim
82	#' @export
83	create_pmap <- function(
84	eventlog,
85	distinct_case = FALSE,
86	distinct_repeated_activities = FALSE,
87	target_categories = NULL,
88	edge_label = c(
89	"amount",
90	"mean_duration",
91	"median_duration",
92	"max_duration",
93	"min_duration"
94	),
95	edge_width = c(
96	"amount",
97	"mean_duration",
98	"median_duration",
99	"max_duration",
100	"min_duration"
101	)
102	) {
103	# Make R Cmd Check happy
104	old_activity <- case_id <- activity <- old_activity <- timestamp <- NULL	12✔
105
106	# clean names
107	eventlog <- eventlog %>%	12✔
108	dplyr::mutate_if(is.factor, as.character) %>%	12✔
109	dplyr::mutate_if(is.character, stringr::str_trim)	12✔
110
111	if (distinct_repeated_activities) {	12✔
112	eventlog <- eventlog %>%	2✔
113	dplyr::group_by(case_id, activity) %>%	2✔
114	dplyr::arrange(timestamp, .by_group = TRUE) %>%	2✔
115	dplyr::rename(old_activity = activity) %>%	2✔
116	dplyr::mutate(activity = paste0(old_activity, " (", 1:n(), ")")) %>%	2✔
117	dplyr::ungroup()	2✔
118
119	if (!"category" %in% colnames(eventlog)) {	2✔
120	# if the `category` is missing,
121	# then use the original `activity` as the `category`
122	eventlog <- eventlog %>% dplyr::rename(category = old_activity)	1✔
123	}
124	}
125
126	nodes <- generate_nodes(eventlog, distinct_case)	12✔
127	edges <- generate_edges(eventlog, distinct_case, target_categories)	12✔
128
129	if (nrow(nodes) == 0) {	12✔
130	stop("Generated graph contains empty node.")	×
131	}
132
133	if (nrow(edges) == 0) {	12✔
134	stop("Generated graph contains no edge.")	×
135	}
136
137	p <- create_pmap_graph(nodes, edges, target_categories, edge_label, edge_width)	12✔
138	return(p)	12✔
139	}

twang2218 / pmap / #93

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous