|
| 1 | +library(jsonlite) |
| 2 | +library(tidyverse) |
| 3 | +library(lubridate) |
| 4 | + |
| 5 | +# path <- "G:/Saját meghajtó/HiFly/Common/Homokozó/nlp-hackathon/data/" |
| 6 | + |
| 7 | +# data_in <- readRDS(paste0(path, "collected/items_225_245_v2.RDS")) |
| 8 | +# ABT_in <- read.csv(paste0(path, "transformed/ABT_base.csv")) |
| 9 | + |
| 10 | +# abt_date <- "2020-07-01" |
| 11 | + |
| 12 | +add_hist_activity <- function(ABT_in, data_in, abt_date){ |
| 13 | + ABT <- ABT_in %>% filter(year_month == abt_date) |
| 14 | + |
| 15 | + data <- data_in %>% |
| 16 | + mutate(time = as.Date(time)) %>% |
| 17 | + filter(time < as.Date(abt_date), |
| 18 | + time > as.Date(abt_date) - as.difftime(183, unit="days"), |
| 19 | + by %in% ABT$by) %>% |
| 20 | + select(by, descendants, score, time) %>% |
| 21 | + replace_na(list("score" = 0, "descendants" = 0)) |
| 22 | + |
| 23 | + activity_30_1 <- data %>% |
| 24 | + filter(as.Date(abt_date) - time < 30) %>% |
| 25 | + group_by(by) %>% |
| 26 | + summarise(max_activity_score_30 = max(score), |
| 27 | + max_activity_desc_30 = max(descendants), |
| 28 | + mean_activity_score_30 = mean(score), |
| 29 | + mean_activity_desc_30 = mean(descendants), |
| 30 | + activity_count_30 = n()) %>% |
| 31 | + ungroup() %>% |
| 32 | + replace(is.na(.), 0) |
| 33 | + |
| 34 | + activity_30_2 <- data %>% |
| 35 | + filter(as.Date(abt_date) - time < 60, |
| 36 | + as.Date(abt_date) - time >= 30) %>% |
| 37 | + group_by(by) %>% |
| 38 | + summarise(max_activity_score_30_2 = max(score), |
| 39 | + max_activity_desc_30_2 = max(descendants), |
| 40 | + mean_activity_score_30_2 = mean(score), |
| 41 | + mean_activity_desc_30_2 = mean(descendants), |
| 42 | + activity_count_30_2 = n()) %>% |
| 43 | + ungroup() %>% |
| 44 | + replace(is.na(.), 0) |
| 45 | + |
| 46 | + activity_60 <- data %>% |
| 47 | + filter(as.Date(abt_date) - time < 60) %>% |
| 48 | + group_by(by) %>% |
| 49 | + summarise(max_activity_score_60 = max(score), |
| 50 | + max_activity_desc_60 = max(descendants), |
| 51 | + mean_activity_score_60 = mean(score), |
| 52 | + mean_activity_desc_60 = mean(descendants), |
| 53 | + activity_count_60 = n()) %>% |
| 54 | + ungroup() %>% |
| 55 | + replace(is.na(.), 0) |
| 56 | + |
| 57 | + activity_60_2 <- data %>% |
| 58 | + filter(as.Date(abt_date) - time < 120, |
| 59 | + as.Date(abt_date) - time >= 60) %>% |
| 60 | + group_by(by) %>% |
| 61 | + summarise(max_activity_score_60_2 = max(score), |
| 62 | + max_activity_desc_60_2 = max(descendants), |
| 63 | + mean_activity_score_60_2 = mean(score), |
| 64 | + mean_activity_desc_60_2 = mean(descendants), |
| 65 | + activity_count_60_2 = n()) %>% |
| 66 | + ungroup() %>% |
| 67 | + replace(is.na(.), 0) |
| 68 | + |
| 69 | + activity_trend_30 <- activity_30_1 %>% |
| 70 | + left_join(activity_30_2) %>% |
| 71 | + replace(is.na(.), 0) %>% |
| 72 | + mutate(max_activity_score_t_30 = max_activity_score_30 - max_activity_score_30_2, |
| 73 | + max_activity_desc_t_30 = max_activity_desc_30 - max_activity_desc_30_2, |
| 74 | + mean_activity_score_t_30 = mean_activity_score_30 - mean_activity_score_30_2, |
| 75 | + mean_activity_desc_t_30 = mean_activity_desc_30 - mean_activity_desc_30_2, |
| 76 | + activity_trend_30 = activity_count_30 - activity_count_30_2) |
| 77 | + |
| 78 | + activity_trend_60 <- activity_60 %>% |
| 79 | + left_join(activity_60_2) %>% |
| 80 | + replace(is.na(.), 0) %>% |
| 81 | + mutate(max_activity_score_t_60 = max_activity_score_60 - max_activity_score_60_2, |
| 82 | + max_activity_desc_t_60 = max_activity_desc_60 - max_activity_desc_60_2, |
| 83 | + mean_activity_score_t_60 = mean_activity_score_60 - mean_activity_score_60_2, |
| 84 | + mean_activity_desc_t_60 = mean_activity_desc_60 - mean_activity_desc_60_2, |
| 85 | + activity_trend_60 = activity_count_60 - activity_count_60_2) |
| 86 | + |
| 87 | + ABT <- ABT %>% |
| 88 | + left_join(activity_trend_30) %>% |
| 89 | + left_join(activity_trend_60) %>% |
| 90 | + replace(is.na(.), 0) |
| 91 | + |
| 92 | + return(ABT) |
| 93 | +} |
0 commit comments