CODEWITHSUNDEEP
rggplot2timestampdata-visualizationhistogram
Emma Vaze
Emma Vaze

Reputation: 19

Timestamp in R plot

The timestamp data is in chr as follows:

posted_at: chr  "2012-01-29 19:48:33" "2012-02-02 15:53:13" "2012-10-24 17:11:40" "2014-07-12 17:00:00" ...
   [1] "2012-01-29 19:48:33" "2012-02-02 15:53:13" "2012-10-24 17:11:40" "2014-07-12 17:00:00" "2014-07-31 08:08:31"
   [6] "2014-07-31 10:48:25" "2014-08-06 09:24:38" "2015-06-16 15:55:28" "2015-06-16 19:56:28" "2015-06-25 17:20:29"
  [11] "2015-06-26 18:28:31"

I tried converting it using strptime() :

tweet_text$posted_at <- strptime(trump_text$posted_at, "%Y-%m-%d %H:%M:%S")

It gets converted as follows:

posted_at: POSIXlt, format: "2012-01-29 19:48:33" "2012-02-02 15:53:13" "2012-10-24 17:11:40" "2014-07-12 17:00:00" ...
   [1] "2012-01-29 19:48:33 AEDT" "2012-02-02 15:53:13 AEDT" "2012-10-24 17:11:40 AEDT" "2014-07-12 17:00:00 AEST"
   [5] "2014-07-31 08:08:31 AEST" "2014-07-31 10:48:25 AEST" "2014-08-06 09:24:38 AEST" "2015-06-16 15:55:28 AEST"
   [9] "2015-06-16 19:56:28 AEST" "2015-06-25 17:20:29 AEST" "2015-06-26 18:28:31 AEST" "2015-07-01 15:58:41 AEST"
  [13] "2015-07-01 18:05:15 AEST

How do I use hist() to plot this timestamp data based on years and count and other different combinations .

Upvotes: 0

Views: 316

Answers (2)

Ronak Shah
Ronak Shah

Reputation: 389055

POSIXlt class already has year component in it which shows number of years since 1900. You can add 1900 to the year and use it in hist.

hist(trump_text$posted_at$year + 1900)

Using ggplot2 you can do :

library(dplyr)
library(ggplot2)

trump_text %>%
  mutate(year = format(posted_at, '%Y')) %>%
  ggplot() + aes(year) + geom_histogram(stat = 'count')

enter image description here

You can customize/update the plot based on your choice.

Upvotes: 1

akrun
akrun

Reputation: 887251

We can take the frequency count with table on the extracted 'Year' from the Datetime object and do a barplot or hist. No external packages used

hist(as.numeric(format(trump_text$posted_at, '%Y')))

Using a reproducible exampe

v1 <- as.POSIXlt(sample(seq(Sys.time(), length.out = 20, by = 'year'), 200, replace = TRUE))
hist(as.numeric(format(v1, '%Y')))

-output enter image description here

Or another option is table with barplot

barplot(table(format(v1, '%Y')))

Or using tidyverse

library(dplyr)
library(lubridate)
library(ggplot2)
tibble(v1 = v1) %>%
     mutate(year = year(v1)) %>%
     ggplot(aes(year)) +
     geom_histogram(stat = 'count')

-output

enter image description here

Upvotes: 1

Related Questions