I am working with Covid data at districts level & trying to plot Cases timeseries for Districts using facet_wrap().
Some States have lots of districts and all of that wont fit so I want to limit the facet_wrap to top N districts.
I have tried to reorder the facet_wrap() by fct_reorder(Districts) but that only reorders & produces plot for all the districts.
Is there a way I can get top N levels of those fct_reorder(Districts) and plot only those top N or if there is any option to control the number of facets in facet_wrap ?
df:
library(tidyverse)
library(lubridate)
file_url <- "https://raw.githubusercontent.com/johnsnow09/covid19-df_stack-code/main/df_districts.csv"
df_districts <- read.csv(url(file_url))
df_districts <- df_districts %>%
mutate(Date = ymd(Date))
I have tried:
df_districts %>%
filter(State == "Rajasthan",
Date != max(Date),
!is.na(Daily_confirmed)) %>%
# group_by(District, Date) %>%
# slice_max(order_by = Daily_confirmed, n = 20) %>%
mutate(District = fct_reorder(District, Daily_confirmed,
.fun = max, .desc = TRUE)) %>%
ggplot(aes(x = Date, y = Daily_confirmed)) +
geom_line(size = 1) +
facet_wrap(~District)
One way is to summarize the table by whatever precedence you need (e.g., sum(Daily_confirmed)), then take the top "n" by that sorted variable.
df_districts %>%
group_by(District) %>%
summarize(daily = sum(Daily_confirmed)) %>%
slice_max(daily, n = 10)
# # A tibble: 10 x 2
# District daily
# <chr> <int>
# 1 Jaipur 99843
# 2 Jodhpur 72443
# 3 Kota 39442
# 4 Alwar 34650
# 5 Udaipur 31297
# 6 Bikaner 26144
# 7 Ajmer 25866
# 8 Bhilwara 19922
# 9 Pali 16589
# 10 Sikar 15031
Using this data, we can left_join the original data back in on District (removing daily first, if desired), and your subset will have just those districts.
out <- df_districts %>%
group_by(District) %>%
summarize(daily = sum(Daily_confirmed)) %>%
slice_max(daily, n = 10) %>%
select(-daily) %>%
left_join(df_districts, by = "District")
out
# # A tibble: 3,660 x 11
# District Date State Confirmed Recovered Deceased Other Tested Daily_confirmed Daily_Recovered Daily_Deceased
# <chr> <date> <chr> <int> <int> <int> <int> <int> <int> <int> <int>
# 1 Jaipur 2021-04-27 Rajasthan 100651 68325 659 0 968783 3289 989 21
# 2 Jaipur 2021-04-26 Rajasthan 97362 67336 638 0 968783 2878 961 11
# 3 Jaipur 2021-04-25 Rajasthan 94484 66375 627 0 968783 3145 648 13
# 4 Jaipur 2021-04-24 Rajasthan 91339 65727 614 0 968783 3260 657 9
# 5 Jaipur 2021-04-23 Rajasthan 88079 65070 605 0 968783 3036 767 13
# 6 Jaipur 2021-04-22 Rajasthan 85043 64303 592 0 968783 2317 658 11
# 7 Jaipur 2021-04-21 Rajasthan 82726 63645 581 0 968783 3101 551 5
# 8 Jaipur 2021-04-20 Rajasthan 79625 63094 576 0 968783 1875 413 9
# 9 Jaipur 2021-04-19 Rajasthan 77750 62681 567 0 968783 2011 528 11
# 10 Jaipur 2021-04-18 Rajasthan 75739 62153 556 0 968783 1963 258 4
# # ... with 3,650 more rows
For comparison,
table(df_districts$District)
# Ajmer Alwar Banswara Baran Barmer Bharatpur Bhilwara Bikaner
# 366 366 366 362 366 366 366 366
# BSF Camp Bundi Chittorgarh Churu Dausa Dholpur Dungarpur Evacuees
# 356 335 366 366 366 366 366 366
# Ganganagar Hanumangarh Italians Jaipur Jaisalmer Jalore Jhalawar Jhunjhunu
# 342 366 366 366 366 356 366 366
# Jodhpur Karauli Kota Nagaur Other State Pali Pratapgarh Rajsamand
# 366 366 366 366 366 366 366 366
# Sawai Madhopur Sikar Sirohi Tonk Udaipur
# 366 366 355 366 366
table(out$District)
# Ajmer Alwar Bhilwara Bikaner Jaipur Jodhpur Kota Pali Sikar Udaipur
# 366 366 366 366 366 366 366 366 366 366