I have a DF named JOB. In that DF i have 4 columns. Person_ID; JOB; FT (full time or part time with values of 1 for full time and 2 for part time) and YEAR. Every person can have only 1 full time job per year in this DF. This is the full time job they got most of their income during the year.
DF
PERSON_ID JOB FT YEAR
1 Analyst 1 2018
1 Analyst 1 2019
1 Analyst 1 2020
2 Coach 1 2018
2 Coach 1 2019
2 Analyst 1 2020
3 Gardener 1 2020
4 Coach 1 2018
4 Coach 1 2019
4 Analyst 1 2020
4 Coach 2 2019
4 Gardener 2 2019
I want to get different frequency in the lines of the following question:
What full time job changes occurred from 2019 and 2020?
I want to look only at changes where FT=1.
I want my end table to look like this
2019 2020 frequency
Analyst Analyst 1
Coach Analyst 2
NA Gardener 1
I want to look at the data so that i can say 2 people moved from they coaching job to analyst job. 1 analyst did not change their job and one person entered the labour market as a gardener.
I tried to fiddle around with the table function but did not even get close to what i wanted. I could not get the YEAR's to go to separate variables.
10 Bonus points if i can do it in base R :)
Thank you for your help
Not pretty but worked:
# split df by year
df_2019 <- df[df$YEAR %in% c(2019) & df$FT == 1, ]
df_2020 <- df[df$YEAR %in% c(2020) & df$FT == 1, ]
# rename Job columns
df_2019$JOB_2019 <- df_2019$JOB
df_2020$JOB_2020 <- df_2020$JOB
# select needed columns
df_2019 <- df_2019[, c("PERSON_ID", "JOB_2019")]
df_2020 <- df_2020[, c("PERSON_ID", "JOB_2020")]
# merge dfs
df2 <- merge(df_2019, df_2020, by = "PERSON_ID", all = TRUE)
df2$frequency <- 1
df2$JOB_2019 <- addNA(df2$JOB_2019)
df2$JOB_2020 <- addNA(df2$JOB_2020)
# aggregate frequency
aggregate(frequency ~ JOB_2019 + JOB_2020, data = df2, FUN = sum, na.action=na.pass)
JOB_2019 JOB_2020 frequency
1 Analyst Analyst 1
2 Coach Analyst 2
3 <NA> Gardener 1
Not R base but worked:
library(dplyr)
library(tidyr)
data %>%
filter(FT==1, YEAR %in% c(2019, 2020)) %>%
group_by(YEAR, JOB, PERSON_ID) %>%
tally() %>%
pivot_wider(names_from = YEAR, values_from = JOB) %>%
select(-PERSON_ID) %>%
group_by(`2019`, `2020`) %>%
summarise(n = n())
`2019` `2020` n
<chr> <chr> <int>
1 Analyst Analyst 1
2 Coach Analyst 2
3 NA Gardener 1
Related
I’ve to transform my dataframe from the current to the new format (see image or structure below). I’ve no idea how I can accomplish that. I want a year for each ID, from 2013-2018 (so each ID has 6 rows, one for every year). The dates are the dates of living on that adress (entry date) and when they left that adress (end date). So each ID and year gives the zipcode and city they lived. The place the ID lived (for each year) should be were they lived the longest that year. I've already set the enddate to 31-12-2018 if they still live there (here showed with NA). Below a picture and the first 3 rows. Hopefully you guys can help me out!
Current format:
ID (1, 1, 2)
ZIPCODE (1234AB, 5678CD, 9012EF)
CITY (NEWYORK, LA, MIAMI)
ENTRY_DATE (2-1-2014, 13-3-2017, 10-11-2011)
END_DATE (13-5-2017, 21-12-2018, 6-9-2017)
New format:
ID (1, 1, 1, 1, 1, 1, 2)
YEAR (2013, 2014, 2015, 2016, 2017, 2018, 2013)
ZIPCODE (NA, 1234AB, 1234AB, 1234AB, 5678CD, 5678CD, 9012EF)
CITY (NA, NEWYORK, NEWYORK, NEWYORK, LA, LA, MIAMI)
See link below
Here is one approach.
First, create date intervals for each location from start to end dates. Using map2 and unnest you will create additional rows for each year.
Since you wish to include the location information where there were the greatest number of days for that calendar year, you could look at overlaps between 2 intervals: one interval is the calendar year, and the second interval is the ENTRY_DATE to END_DATE. For each year, you can filter by max(WEEKS) (or to ensure a single address per year, arrange in descending order by WEEKS and slice(1) --- or with latest tidyr consider slice_max). This will keep the row where there is the greatest number of weeks duration overlap between intervals.
The final complete will ensure you have rows for all years between 2013-2018.
library(tidyverse)
library(lubridate)
df %>%
mutate(ENTRY_END_INT = interval(ENTRY_DATE, END_DATE),
YEAR = map2(year(ENTRY_DATE), year(END_DATE), seq)) %>%
unnest(YEAR) %>%
mutate(YEAR_INT = interval(as.Date(paste0(YEAR, '-01-01')), as.Date(paste0(YEAR, '-12-31'))),
WEEKS = as.duration(intersect(ENTRY_END_INT, YEAR_INT))) %>%
group_by(ID, YEAR) %>%
arrange(desc(WEEKS)) %>%
slice(1) %>%
group_by(ID) %>%
complete(YEAR = seq(2013, 2018, 1)) %>%
arrange(ID, YEAR) %>%
select(-c(ENTRY_DATE, END_DATE, ENTRY_END_INT, YEAR_INT, WEEKS))
Output
# A tibble: 14 x 4
# Groups: ID [2]
ID YEAR ZIPCODE CITY
<dbl> <dbl> <chr> <chr>
1 1 2013 NA NA
2 1 2014 1234AB NEWYORK
3 1 2015 1234AB NEWYORK
4 1 2016 1234AB NEWYORK
5 1 2017 5678CD LA
6 1 2018 5678CD LA
7 2 2011 9012EF MIAMI
8 2 2012 9012EF MIAMI
9 2 2013 9012EF MIAMI
10 2 2014 9012EF MIAMI
11 2 2015 9012EF MIAMI
12 2 2016 9012EF MIAMI
13 2 2017 9012EF MIAMI
14 2 2018 NA NA
Data
df <- structure(list(ID = c(1, 1, 2), ZIPCODE = c("1234AB", "5678CD",
"9012EF"), CITY = c("NEWYORK", "LA", "MIAMI"), ENTRY_DATE = structure(c(16072,
17238, 15288), class = "Date"), END_DATE = structure(c(17299,
17896, 17415), class = "Date")), class = "data.frame", row.names = c(NA,
-3L))
I have a dataframe with id variable name. I'm trying to figure out a way to transpose each variable in the dataframe by name.
My current df is below:
name jobtitle companyname datesemployed empduration joblocation jobdescrip
1 David… Project… EOS IT Man… Aug 2018 – P… 1 yr 9 mos San Franci… Coordinati…
2 David… Technic… Options Te… Sep 2017 – J… 5 mos Belfast, U… Working wi…
3 David… Data An… NA Jan 2018 – J… 6 mos Belfast, U… Working wi…
However, I'd like a dataframe in which there is only one row for name, and every observation for name becomes its own column, like below:
name jobtitle_1 companyname_1 datesemployed_1 empduration_1 joblocation_1 jobdescrip_1 job_title2 companyname_2 datesemployed_2 empduration_2 joblocation_2 jobdescrip_2
1 David… Project… EOS IT Man… Aug 2018 – P… 1 yr 9 mos San Franci… Coordinati… Technic… Options Te… Sep 2017 – J… 5 mos Belfast, U… Working wi…
I have used commands like gather_by and melt in the past to reshape from long to wide, but in this case, I'm not sure how to apply it, since every observation for the id variable will need to become its own column.
It sounds like you are looking for gather and pivot_wider.
I used my own sample data with two names:
df <- tibble(name = c('David', 'David', 'David', 'Bill', 'Bill'),
jobtitle = c('PM', 'TPM', 'Analyst', 'Dev', 'Eng'),
companyname = c('EOS', 'Options', NA, 'Microsoft', 'Nintendo'))
First add an index column to distinguish the different positions for each name.
indexed <- df %>%
group_by(name) %>%
mutate(.index = row_number())
indexed
# name jobtitle companyname .index
# <chr> <chr> <chr> <int>
# 1 David PM EOS 1
# 2 David TPM Options 2
# 3 David Analyst NA 3
# 4 Bill Dev Microsoft 1
# 5 Bill Eng Nintendo 2
Then it is possible to use gather to get a long form, with one value per row.
gathered <- indexed %>% gather('var', 'val', -c(name, .index))
gathered
# name .index var val
# <chr> <int> <chr> <chr>
# 1 David 1 jobtitle PM
# 2 David 2 jobtitle TPM
# 3 David 3 jobtitle Analyst
# 4 Bill 1 jobtitle Dev
# 5 Bill 2 jobtitle Eng
# 6 David 1 companyname EOS
# 7 David 2 companyname Options
# 8 David 3 companyname NA
# 9 Bill 1 companyname Microsoft
# 10 Bill 2 companyname Nintendo
Now pivot_wider can be used to create a column for each variable and index.
gathered %>% pivot_wider(names_from = c(var, .index), values_from = val)
# name jobtitle_1 jobtitle_2 jobtitle_3 companyname_1 companyname_2 companyname_3
# <chr> <chr> <chr> <chr> <chr> <chr> <chr>
# 1 David PM TPM Analyst EOS Options NA
# 2 Bill Dev Eng NA Microsoft Nintendo NA
Get the data in long format, create a unique column identifier and get it back to wide format.
library(dplyr)
library(tidyr)
df %>%
pivot_longer(cols = -name, names_to = 'col') %>%
group_by(name, col) %>%
mutate(row = row_number()) %>%
pivot_wider(names_from = c(col, row), values_from = value)
I found that dplyr is speedy and simple for aggregate and summarise data. But I can't find out how to solve the following problem with dplyr.
Given these data frames:
df_2017 <- data.frame(
expand.grid(1:195,1:65,1:39),
value = sample(1:1000000,(195*65*39)),
period = rep("2017",(195*65*39)),
stringsAsFactors = F
)
df_2017 <- df_2017[sample(1:(195*65*39),450000),]
names(df_2017) <- c("company", "product", "acc_concept", "value", "period")
df_2017$company <- as.character(df_2017$company)
df_2017$product <- as.character(df_2017$product)
df_2017$acc_concept <- as.character(df_2017$acc_concept)
df_2017$value <- as.numeric(df_2017$value)
ratio_df <- data.frame(concept=c("numerator","numerator","numerator","denom", "denom", "denom","name"),
ratio1=c("1","","","4","","","Sales over Assets"),
ratio2=c("1","","","5","6","","Sales over Expenses A + B"), stringsAsFactors = F)
where the columns in df_2017 are:
company = This is a categorical variable with companies from 1 to 195
product = This is a categorical, with home apliance products from 1 to 65. For example, 1 could be equal to irons, 2 to television, etc
acc_concept = This is a categorical variable with accounting concepts from 1 to 39. For example, 1 would be equal to "Sales", 2 to "Total Expenses", 3 to Returns", 4 to "Assets, etc
value = This is a numeric variable, with USD from 1 to 100.000.000
period = Categorical variable. Always 2017
As the expand.grid implies, the combinations of company - product - acc_concept are never duplicated, but, It could happen that certains subjects have not every company - product - acc_concept combinations. That's why the code line "df_2017 <- df_2017[sample(1:195*65*39),450000),]", and that's why the output could turn out into NA (see below).
And where the columns in ratio_df are:
Concept = which acc_concept corresponds to numerator, which one to
denominator, and which is name of the ratio
ratio1 = acc_concept and name for ratio1
ratio2 = acc_concept and name for ratio2
I want to calculate 2 ratios (ratio_df) between acc_concept, for each product within each company.
For example:
I take the first ratio "acc_concepts" and "name" from ratio_df:
num_acc_concept <- ratio_df[ratio_df$concept == "numerator", 2]
denom_acc_concept <- ratio_df[ratio_df$concept == "denom", 2]
ratio_name <- ratio_df[ratio_df$concept == "name", 2]
Then I calculate the ratio for one product of one company, just to show you want i want to do:
ratio1_value <- sum(df_2017[df_2017$company == 1 & df_2017$product == 1 & df_2017$acc_concept %in% num_acc_concept, 4]) / sum(df_2017[df_2017$company == 1 & df_2017$product == 1 & df_2017$acc_concept %in% denom_acc_concept, 4])
Output:
output <- data.frame(Company="1", Product="1", desc_ratio=ratio_name, ratio_value = ratio1_value, stringsAsFactors = F)
As i said before i want to do this for each product within each company
The output data.frame could be something like this (ratios aren't the true ones because i haven't done the calculations yet):
company product desc_ratio ratio_value
1 1 Sales over Assets 0.9303675
1 2 Sales over Assets 1.30
1 3 Sales over Assets Nan
1 4 Sales over Assets Inf
1 5 Sales over Assets 2.32
1 6 Sales over Assets NA
.
.
.
1 1 Sales over Expenses A + B 3.25
.
.
.
2 1 Sales over Assets 0.256
and so on...
NaN when ratio is 0 / 0
Inf when ratio is number / 0
NA when there is no data for certain company and product.
I hope i have made myself clear this time :)
Is there any way to solve this row problem with dplyr? Should I cast the df_2017 for mutating? In this case, which is the best way for casting?
Any help would be welcome!
This is one way of doing it. At the end I timed the code on all of your records.
First create a function to create all the ratios. Do note, this function is only useful inside the dplyr code.
ratio <- function(data){
result <- data.frame(desc_ratio = rep(NA, ncol(ratio_df) -1), ratio_value = rep(NA, ncol(ratio_df) -1))
for(i in 2:ncol(ratio_df)){
num <- ratio_df[ratio_df$concept == "numerator", i]
denom <- ratio_df[ratio_df$concept == "denom", i]
result$desc_ratio[i-1] <- ratio_df[ratio_df$concept == "name", i]
result$ratio_value[i-1] <- sum(ifelse(data$acc_concept %in% num, data$value, 0)) / sum(ifelse(data$acc_concept %in% denom, data$value, 0))
}
return(result)
}
Using dplyr, tidyr and purrr to put everything together. First group by the data, nest the data needed for the function, run the function with a mutate on the nested data. Drop the not needed nested data and unnest to get your wanted output. I leave the sorting up to you.
library(dplyr)
library(purrr)
library(tidyr)
output <- df_2017 %>%
group_by(company, product, period) %>%
nest() %>%
mutate(ratios = map(data, ratio)) %>%
select(-data) %>%
unnest
output
# A tibble: 25,350 x 5
company product period desc_ratio ratio_value
<chr> <chr> <chr> <chr> <dbl>
1 103 2 2017 Sales over Assets 0.733
2 103 2 2017 Sales over Expenses A + B 0.219
3 26 26 2017 Sales over Assets 0.954
4 26 26 2017 Sales over Expenses A + B 1.01
5 85 59 2017 Sales over Assets 4.14
6 85 59 2017 Sales over Expenses A + B 1.83
7 186 38 2017 Sales over Assets 7.85
8 186 38 2017 Sales over Expenses A + B 0.722
9 51 25 2017 Sales over Assets 2.34
10 51 25 2017 Sales over Expenses A + B 0.627
# ... with 25,340 more rows
Time it took to run this code on my machine measured with system.time:
user system elapsed
6.75 0.00 6.81
I have a dataset of patents, where I have recorded 1) the month and year associated with a patent renewal and 2) whether the patent holder chose to pay the patent fee or let the patent lapse. So
patentid fee1date fee1paid fee2date fee2paid
1 May 2010 True May 2013 False
2 May 2010 True April 2014 True
What I want to do is develop a count of the number of renewals by month and by year, as well as the number of abandoned patents, as follows:
date renewed lapsed
May 2010 2 0
How might I count the data that I have now? Thank you!
EDIT: The key point is to aggregate these across different columns. The issue that I am running into now is that when I try using the count library, it treats 2 renewals in May 2010 as two separate values.
Using dplyr
require(tidyr)
require(dplyr)
data %>% gather(year,value, -Patent.ID) %>%
separate('year',c('Fee','N','Act')) %>%
spread(Act,value) %>%
unite(Fee, Fee,N, sep = '.') %>%
group_by(Date) %>%
summarise(R=sum(Paid=='True'), NotR=sum(Paid=='False'))
# A tibble: 3 x 3
Date R NotR
<chr> <int> <int>
1 April 2014 1 0
2 May 2010 2 0
3 May 2013 0 1
Data
data <- read.table(text="
'Patent ID' 'Fee 1 Date' 'Fee 1 Paid' 'Fee 2 Date' 'Fee 2 Paid'
1 'May 2010' True 'May 2013' False
2 'May 2010' True 'April 2014' True
",header=T, stringsAsFactors = F)
Problem Description :
I am trying to calculate the recency , based on , what is the most recent value in Year column where the target achieved indicator was equal to 1 and in case the indicator column has 0 as the only available value for the Salesman + Year key, choose the minimum year in that case
Data:
Salesman_ID Year Yearly_Targets_Achieved_Indicator
1 AA-5468 2012 1
2 AA-5468 2013 0
3 AA-5468 2014 0
4 AA-5468 2015 0
5 AA-5468 2016 1
6 AL-3791 2012 1
7 AL-3791 2013 1
8 AL-3791 2014 0
9 AL-3893 2015 0
10 AL-3893 2016 0
Expected Output:
Salesman_ID Year Yearly_Targets_Achieved_Indicator
<chr> <dbl> <dbl>
1 AA-5468 2016 1
2 AA-3791 2013 1
9 AL-3893 2015 0
Using the package tidyverse I suggest you the following code:
library(tidyverse)
Prashant_df <- data.frame(
c("AA-5468","AA-5468","AA-5468","AA-5468","AA-5468","AL-3791","AL-3791","AL-3791","AL-3893","AL-3893"),
c(2012,2013,2014,2015,2016,2012,2013,2014,2015,2016),
c(1,0,0,0,1,1,1,0,0,0)
)
names(Prashant_df) <- c("Salesman_ID","Year","Yearly_Targets_Achieved_Indicator")
Prashant_df <- Prashant_df %>%
group_by(Salesman_ID) %>%
mutate(Year_target=case_when(
Yearly_Targets_Achieved_Indicator==1 ~ max(Year),
Yearly_Targets_Achieved_Indicator==0 ~ min(Year)
))
Prashant_df_collapsed <- Prashant_df %>%
group_by(Salesman_ID) %>%
summarise(Year=max(Year_target),
Yearly_Targets_Achieved_Indicator=max(Yearly_Targets_Achieved_Indicator))
You can store both maximum and minimum year for each salesman, and the maximum of your binary variable.
newdf = df %>% group_by(Salesman_ID) %>% summarise(
maximum = max(Year),
minimum = min(Year),
maxInd = max(Yearly_Targets_Achieved_Indicator))
From this you can pretty much construct your resulting variable.
Using Base R:
c(by(dat,dat[1],function(x)if(all(x[,3]==0)) x[1,2] else max(x[which(x[,3]==1),2])))
AA-5468 AL-3791 AL-3893
2016 2013 2015
This code is kind of a messy but produces the desired output: Here is the explanation:
first groupby salesman_id, then for that specific group check whether all the indicators are zero, if yes, return the first year. else, look for the latest/maximum year among those which the indicators are 1