r - dplyr mutate based on two vectors (calculating a 3rd column based on grouping from 2 columns) -

this question has answer here:

• mean per group in data.frame [duplicate] 8 answers

i have data includes date, user_id, steps, heart rate.

i trying use dplyr mutate give me column have average heart rate per day per user_id. there multiple recordings heart rate throughout day per user. (note: appending column b/c want keep information other columns such 'steps')

code generate data sample

df7 <- data.frame(  date=c('2016-11-01','2016-11-01','2016-11-01','2016-11-01','2016-11-02','2016-11-02','2016-11-02','2016-11-02'),                users_user_id=c(6,6,7,7,6,6,7,7),                steps=c(500,2000,500,2000,600,3000,600,3000),                avg_heart_rate=c(70,80,70,80,80,90,80,90)) df7$date <- as.date(df7$date) 

ideally this

date        users_user_id   steps   average_heart_rate  day_avg_hr 2016-11-01  6               500     70                  75 2016-11-01  6               2000    80                  75 2016-11-01  7               500     70                  75 2016-11-01  7               2000    80                  75 2016-11-02  6               600     80                  85 2016-11-02  6               3000    90                  85 2016-11-02  7               600     80                  85 2016-11-02  7               3000    90                  85 

first summarise join results.

new.df <- df7 %>%   group_by(., date, users_user_id) %>%   summarise(., day_avg_hr = mean(avg_heart_rate)) %>%   ungroup() %>%   left_join(df7, .)