5.2 Daily downloads of R
In this section, we studied the daily downloads of R language itself from 2012-10-01 to 2021-06-12. The data was obtained from the cranlogs package (Csárdi 2019) as well.
Finding 1: There is an upward trend in the number of downloads, also with an increasing variance.
Figure 5.7 shows the download trend of R on CRAN over the period of time, from 2012-10-01 to 2021-06-12. An upward trend shows, with an increasing variance in download counts, which is resulted from the growth of R users. As for significant spikes, they are probably quite similar to the situation of R-packages, which is due to repeated downloads, weekly calendar effect or server test issue.
Figure 5.7: The download trend of R on CRAN, from 2012-10-01 to 2021-06-12.
Finding 2: The most used operation system for R users turns out to be Windows OS.
Figure 5.8 shows the comparison of operation systems for R users. The number of Windows-user ranks first, followed by MacOS and SRC. Microsoft has traditionally dominated the desktop and laptop market. And Microsoft officially claims there are 400 million active users of Windows 10 itself, while Apple revealed that there are only around 100 million active Mac users, up to now.(Warren 2017)
Figure 5.8: The comparison of operation system (OS) for R users, from 2012-10-01 to 2021-06-12.
Finding 3: The most popular version of R is 3.2.1
Table 5.8 shows the first 10 numbers of downloads for different versions of R. The most downloaded one is 3.2.1 with 4526 times, followed by 3.3.3 with 4305 times and 3.1.2 with 3958 times. Most users are likely to install the latest version of R to get the newest update information.
| R_version | count |
|---|---|
| 3.2.1 | 4526 |
| 3.3.3 | 4305 |
| 3.1.2 | 3958 |
| 3.2.2 | 3781 |
| 3.4.0 | 3585 |
| 3.1.1 | 3458 |
| 3.5.1 | 3450 |
| 3.4.3 | 3409 |
| 3.4.1 | 3398 |
| 3.3.0 | 3362 |
In conclusion, the daily download trends of R are quite similar to those of R-packages, and they are both significantly correlated to weekly calendar as well as unusual downloads from CRAN mirrors. In general, they both tend to increase with the growing of R users.