使用 R 中数据帧中的条件对日期数据进行分组

Question

I have a csv file with several variables, as illustrated below (just for example):

Region    crop      product    date_periode

A         aaaa      bilon      2016052q

A         aaaa      mailon     2016021q 

B         cccc      drox       2016042q

A         cccc      marob      2015081q

C         dddd      salon      2016062q

C         dddd      dilon      2016071q

D         aaaa      daxon      2015032q

D         aaaa      bayon      2016042q

the dates are periods : 20170502q : the seconde half of mai 2017 and i want to do a grouping individuals per crop and region such as every time the number of individuals for a date for a crop in a region is less than 5% of the total number of individuals with the same date for a crop in a region this date is grouped with the adjacent date (the date, in this case, can become 2016062q-2016071q if we group the two periods for example), and every time for each culture in each region. if we have this table for exemple :

region    crop       date         Numbre of ID    % of ID

A         aaaa    20170201q         1             1

A         aaaa    20170202q        44            48

A         aaaa    20170301q        30            33

A         aaaa    20170302q        14            15

A         aaaa    20170401q         1             1

A         aaaa    20170402q         1             1

A         aaaa    20170601q         1             1

i want to arrive to have this one after analysing

region      crop      date                    Number of ID      % of ID

A           aaaa      20170201q-20170202q         45              49

A           aaaa      20170301q                   30              33

A           aaaa      20170302q-20170601q         17              18 

I don't know if I'm clear enough but I'm here if you have any questions above, thank you in advance

Answer 1

Using tidyverse we can do this using:

df %>% 
group_by(Region, crop, date_periode) %>% 
summarise(number = n_distinct(product)) %>% 
ungroup() %>% 
left_join(  df %>% 
            group_by(Region, crop) %>% 
            summarise(number_t = n_distinct(product)) %>% 
            ungroup(), by = c("Region", "crop")) %>% 
mutate(Percent = number/number_t)

I think this is what you are getting at? I am assuming Number is the total distinct product .