CODEWITHSUNDEEP
rdplyrlubridate
Brigadeiro
Brigadeiro

Reputation: 2945

Filtering dataset for a specific date with lubridate and dplyr

I am trying to filter my tibble for a specific date. If I try >= that date & <= that same date, the code results in 0 rows being returned. I have identified that it is the <= that returns FALSE for rows that I believe should be TRUE. Below is a working example. I am only interesting in tidyverse and lubridate related responses - please do not post answers using Base R or other packages (I know how to solve the problem in these other ways.

Data

library(dplyr)
library(lubridate)

d <- structure(list(Date = structure(c(1572679159, 1572703322, 1572710363, 1572722578,
                                       1572727765, 1572728767, 1572730032),
                                     class = c("POSIXct", "POSIXt"),
                                     tzone = "UTC", label = c(EndDate = "EndDate")),
                    ID = 1:7), class = c("spec_tbl_df", "tbl_df", "tbl", "data.frame"),
               row.names = c(NA, -7L))

When I try the following, it returns a tibble with 0 rows:

d %>%
  filter(Date >= as_date("2019-11-02"),
         Date <= as_date("2019-11-02"))

When I try this, however, it works. What am I missing? The problem is that, I would like Date <= as_date("2019-11-03") to return TRUE for all '2019-11-03' submissions (I actually have those in my dataset), but it does not.

d %>%
  filter(Date >= as_date("2019-11-02"),
         Date <= as_date("2019-11-03"))

Upvotes: 3

Views: 5850

Answers (2)

Sang won kim
Sang won kim

Reputation: 522

%in% mean to determine whether a particular value is included in a vector operator. So you can use %in% .

d %>%
  filter(Date >= as.Date("2019-11-02") %in%
           (Date <= as_date("2019-11-03")))
# A tibble: 7 x 2
  Date                   ID
  <dttm>              <int>
1 2019-11-02 07:19:19     1
2 2019-11-02 14:02:02     2
3 2019-11-02 15:59:23     3
4 2019-11-02 19:22:58     4
5 2019-11-02 20:49:25     5
6 2019-11-02 21:06:07     6
7 2019-11-02 21:27:12     7

Upvotes: -2

Ronak Shah
Ronak Shah

Reputation: 388982

You need to convert the datetime object to date class and then compare. Also use & to include multiple conditions to check.

library(dplyr)
library(lubridate)

d %>%
  filter(as_date(Date) >= as_date("2019-11-02") & 
         as_date(Date) <= as_date("2019-11-02"))

#  Date                   ID
#  <dttm>              <int>
#1 2019-11-02 07:19:19     1
#2 2019-11-02 14:02:02     2
#3 2019-11-02 15:59:23     3
#4 2019-11-02 19:22:58     4
#5 2019-11-02 20:49:25     5
#6 2019-11-02 21:06:07     6
#7 2019-11-02 21:27:12     7

Moreover, we can also use between

d %>%
  filter(between(as_date(Date), as_date("2019-11-02"), as_date("2019-11-02")))

Upvotes: 6

Related Questions