Please find the input data and expected output as screenshot below:
However, the current plot with the below code:
I feel, I made it too complicated. But I shared input data and expected data along with struggled code along the way. Could you please help us
Mainly there are 2 issues.
1. If mutate is used, undesired stripes appear on the plot
Summarize used, then it is not adding to 100%
2. How can we extract the top contributors
Both have been tried by us but stuck somewhere
# Input data
df <- tibble(
country = c(rep(c("India","USA","Germany","Africa"), each = 8)),
type = c("sms","Other","whatsapp","web","online","shiny","whatsapp","whatsapp",
"sms","sms","sms","web","web","Other","online","whatsapp",
"sms","Other","whatsapp","shiny","online","shiny","whatsapp","whatsapp",
"sms","sms","sms","shiny","online","Other","online","Other"
),
cust = rep(c("google","Apple","wallmart","pg"),8),
quantity = c(10,20,30,40,50,60,70,80,
90,100,15,25,35,45,55,65,
75,85,95,105,10,15,20,25,
30,35,40,45,50,55,60,65)
)
# Without Customer
df %>%
group_by(country,type) %>%
summarise(kpi_wo_cust = sum(quantity)) %>%
ungroup() -> df_wo_cust
# With Customer
df %>%
group_by(country,type,cust) %>%
summarise(kpi_cust = sum(quantity)) %>%
ungroup() -> df_cust
df_combo <- left_join(df_cust, df_wo_cust, by = c("country","type"))
df_combo %>% glimpse()
# Aggregated data for certain KPIs for final plot
df_aggr <- df_combo %>%
group_by(country,type) %>%
mutate(kpi_cust_total = sum(kpi_cust),
per_kpi_cust = 100 * (kpi_cust/kpi_cust_total)) %>%
group_by(country) %>%
# In order to except from repeated counting, selecting unique()
mutate(kpi_cust_uniq_total = sum(kpi_cust) %>% unique(),
per_unq_kpi_cust = 100 * (kpi_cust/kpi_cust_uniq_total) %>% round(4))
#
plt = df_aggr %>% ungroup() %>%#glimpse()
# In order to obtain theTop 2 customers (Major contributor) within country and type
# However, if this code is used, there is an error
# group_by(country, type) %>%
# nest() %>%
# mutate(top_cust = purrr::map_chr(data, function(x){
# x %>% arrange(desc(per_kpi_cust)) %>%
# top_n(2,per_kpi_cust) %>%
# summarise(Cust = paste(cust,round(per_kpi_cust,2), collapse = "<br>")) %>%
# pull(cust)
# })#,data = NULL
# ) %>%
# unnest(cols = data) %>%
group_by(country, type) %>%
# If mutate is used, undesired stripes appear on the plot
# Summarize used, then it is not adding to 100%
mutate(avg_kpi_cust = per_unq_kpi_cust %>% mean()) %>%
#summarise(avg_kpi_cust = per_unq_kpi_cust %>% mean()) %>%
ggplot(aes(x = country,
y = avg_kpi_cust,
fill = type,
text = paste('<br>proportion: ', round(avg_kpi_cust,2), "%",
"<br>country:",country
))) +
geom_bar(stat = "identity"#, position=position_dodge()
) +
coord_flip() +
theme_bw()
ggplotly(plt)
The key was to use distinct() after mutate() instead of summarise()
Also, mean() was the wrong function used earlier instead of sum() which had resulted in incomplete barplot.
library(tidyverse)
library(plotly)
# Input data
df <- tibble(
country = c(rep(c("India","USA","Germany","Africa"), each = 8)),
type = c("sms","Other","whatsapp","web","online","shiny","whatsapp","whatsapp",
"sms","sms","sms","web","web","Other","online","whatsapp",
"sms","Other","whatsapp","shiny","online","shiny","whatsapp","whatsapp",
"sms","sms","sms","shiny","online","Other","online","Other"
),
cust = rep(c("google","Apple","wallmart","pg"),8),
quantity = c(10,20,30,40,50,60,70,80,
90,100,15,25,35,45,55,65,
75,85,95,105,10,15,20,25,
30,35,40,45,50,55,60,65)
)
# Without Customer
df %>%
group_by(country,type) %>%
summarise(kpi_wo_cust = sum(quantity)) %>%
ungroup() -> df_wo_cust
# With Customer
df %>%
group_by(country,type,cust) %>%
summarise(kpi_cust = sum(quantity)) %>%
ungroup() -> df_cust
df_combo <- left_join(df_cust, df_wo_cust, by = c("country","type"))
df_combo %>% glimpse()
# Aggregated data for certain KPIs for final plot
df_aggr <- df_combo %>%
group_by(country,type) %>%
mutate(kpi_cust_total = sum(kpi_cust),
per_kpi_cust = 100 * (kpi_cust/kpi_cust_total)) %>%
group_by(country) %>%
# In order to except from repeated counting, selecting unique()
mutate(kpi_cust_uniq_total = sum(kpi_cust) %>% unique(),
per_unq_kpi_cust = 100 * (kpi_cust/kpi_cust_uniq_total) %>% round(4))
plt = df_aggr %>% ungroup() %>%
# In order to diplay Top 2 customers (Major contributor) within country and type
group_by(country, type) %>%
nest() %>%
mutate(top_cust = purrr::map_chr(data, function(x){
x %>% arrange(desc(per_kpi_cust)) %>%
top_n(2,per_kpi_cust) %>%
summarise(Cust = paste(cust,round(per_kpi_cust,2), collapse = "<br>")) %>%
pull(Cust)
})) %>%
unnest(cols = data) %>%
group_by(country, type) %>%
# If mutate is used, undesired stripes appear on the plot
# Summarize used, then it is not adding to 100%.
# So distinct was used
mutate(avg_kpi_cust = per_unq_kpi_cust %>% sum()) %>%
ungroup() %>%
distinct(country, type, .keep_all = T) %>%
ggplot(aes(x = country,
y = avg_kpi_cust,
fill = type,
text = top_cust
)) +
geom_bar(stat = "identity") +
coord_flip() +
theme_bw()
ggplotly(plt, tooltip = "text")
I have the following script. Option 1 uses a long format and group_by to identify the first step of many where the status equals 0.
Another option (2) is to use apply to calculate this value for each row, and then transform the data to a long format.
The firs option does not scale well. The second does, but I was unable to get it into a dplyr pipe. I tried to solve this with purrr but did not succeeed.
Questions:
Why does the first option not scale well?
How can I transform the second option in a dplyr pipe?
require(dplyr)
require(tidyr)
require(ggplot2)
set.seed(314)
# example data
dat <- as.data.frame(matrix(sample(c(0,1),
size = 9000000,
replace = TRUE,
prob = c(5,95)),
ncol = 9))
names(dat) <- paste("step",1:9, sep="_")
steps <- dat %>% select(starts_with("step_")) %>% names()
# option 1 is slow
dat.cum <- dat %>%
mutate(id = row_number()) %>%
gather(step, status,-id) %>%
group_by(id) %>%
mutate(drop = min(if_else(status==0,match(step, steps),99L))) %>%
mutate(status = if_else(match(step, steps)>=drop,0,1))
ggplot(dat.cum, aes(x = step, fill = factor(status))) +
geom_bar()
# option 2 is faster
dat$drop <- apply(dat,1,function(x) min(which(x==0),99))
dat.cum <- dat %>%
gather(step,status,-drop) %>%
mutate(status = if_else(match(step,steps)>=drop,0,1))
ggplot(dat.cum, aes(x = step, fill = factor(status))) +
geom_bar()
If you would like to map along rows you could do:
dat %>%
mutate(drop2 = map_int(seq_len(nrow(dat)), ~ min(which(dat[.x, ] == 0L), 99L)))
It could be that "gathering and grouping" is faster than Looping:
dat %>%
as_tibble() %>%
select(starts_with("step_")) %>%
mutate(row_nr = row_number()) %>%
gather(key = "col", value = "value", -row_nr) %>%
arrange(row_nr, col) %>%
group_by(row_nr) %>%
mutate(col_index = row_number()) %>%
filter(value == 0) %>%
summarise(drop3 = min(col_index)) %>%
ungroup() %>%
right_join(dat %>%
mutate(row_nr = row_number()),
by = "row_nr") %>%
mutate(drop3 = if_else(is.na(drop3), 99, drop3))
Hi I have a dataframe in wide format that is grouped by Site. Each column represents the abundance a different species(85 total). I am trying to summarize the dataframe to calculate the total number of individuals regardless of species in my data.
df.totals<- df %>% group_by(Site) %>% summarize (total = sum(6:91))
We can gather to 'long' format and then do the sum
library(tidyverse)
df %>%
select(Site, 6:91) %>%
rownames_to_column("rn") %>%
gather(key, val, - Site, -rn) %>%
group_by(Site, rn) %>%
summarise(total = sum(val))
or another option without gathering would be
df %>%
select(Site, 6:91) %>%
transmute(vs, Sum = reduce(.[2:ncol(.)], `+`)) %>%
group_by(Site) %>%
summarise(Sum = sum(Sum))
Using a reproducible example with mtcars
mtcars %>%
select(vs, 4:6) %>%
transmute(vs, Sum = reduce(.[2:ncol(.)], `+`)) %>%
group_by(vs) %>%
summarise(Sum = sum(Sum))
I have generated this summary table based on the df below.
set.seed(1)
df <- data.frame(rep(
sample(c(2012,2016),10, replace = T)),
sample(c('Treat','Control'),10,replace = T),
runif(10,0,1),
runif(10,0,1),
runif(10,0,1))
colnames(df) <- c('Year','Group','V1','V2','V3')
summary.table = df %>%
group_by(Year, Group) %>%
group_by(N = n(), add = TRUE) %>%
summarise_all(funs(sd,median)) %>%
ungroup %>%
mutate(Year = ifelse(duplicated(Year),"",Year))
Is there a way I could display the values related to the median columns as percentages?
I did not know how to use mutate() and scales::percent() for only a subset of columns (I dont want to do it individually, since there will be more columns in the original dataset, making this procedure not practical enough.
What should I have done instead if I wanted to mutate according to a subset of rows?
Thank you
EDIT:
And if it was like this?
summary.table = df %>%
group_by(Year, Group) %>%
summarise_all(funs(median,sd)) %>%
gather(key, value, -Year, -Group) %>%
separate(key, into=c("var", "stat")) %>%
unite(stat_Group, stat, Group) %>%
spread(stat_Group, value) %>%
ungroup %>%
mutate(Year = ifelse(duplicated(Year),"",Year))
We need to use the percent wrapped on median
summary.table <- df %>%
group_by(Year, Group) %>%
group_by(N = n(), add = TRUE) %>%
summarise_all(funs(sd=sd(.),median=scales::percent(median(.)))) %>%
ungroup %>%
mutate(Year = ifelse(duplicated(Year),"",Year))