DFCI Introduction to R and RStudio

Transcript

1. Introduction to R/RStudio
   Patrick Kimes, PhD
   Postdoctoral Fellow
   Dana-Farber Cancer Institute
   Harvard TH Chan School of Public Health
   Top Ten Seminars
   October 1, 2019
   

   View Slide

2. October 1, 2019
   October 22, 2019
   November 12, 2019
   December 10, 2019
   January 21, 2020
   February 11, 2020
   March 17, 2020
   April 14, 2020
   May 19, 2020
   June 16, 2020
   Introduction to R and RStudio
   Data visualization with ggplot2
   Data visualization principles and plots to avoid
   Design of Clinical Trials Basics
   Correlation: you are probably using it wrong
   How to detect and deal with batch effects
   Brief introduction to machine learning
   Culprits of the reproducibility crisis: multiple testing and p-hacking
   Experimental Design: How many size and should I pool?
   Detecting differentially expressed genes with RNA-seq
   top ten seminars
   in data science
   

   View Slide

3. R and RStudio?
   

   View Slide

4. R and RStudio?
   programming
   language
   R
   think Java, C,
   C++, Python, …
   

   View Slide

5. RStudio
   programming
   language
   R and RStudio?
   software to make data
   analysis with R easier
   

   View Slide

6. RStudio
   programming
   language
   R and RStudio?
   programming
   language
   R
   software to make data
   analysis with R easier
   

   View Slide

7. View Slide

8. RStudio
   engine
   R
   dashboard
   

   View Slide

9. why
   R and RStudio?
   

   View Slide

10. what about SAS?
    why
    R and RStudio?
    

    View Slide

11. what about SAS?
    • R is free, open source
    • R is the home of new methods
    • R has a large, active community
    • R is highly interoperable, extensible
    why
    R and RStudio?
    

    View Slide

12. what about Python?
    why
    R and RStudio?
    

    View Slide

13. what about Python?
    • Good question! Up to you!
    • R is arguably easier to learn
    • R has more statistical tools
    • R makes exploration and visualization easier
    why
    R and RStudio?
    

    View Slide

14. why
    R and RStudio?
    

    View Slide

15. it gets you to
    the data fast!
    why
    R and RStudio?
    

    View Slide

16. it gets you to
    the data fast!
    and that’s fun!
    https://twitter.com/avogado6/status/1165595520967954432
    

    View Slide

17. who already has
    R / RStudio installed?
    

    View Slide

18. how to install
    R and RStudio
    

    View Slide

19. View Slide

20. how to install
    R and RStudio
    1. Search “R”, 

    Search “RStudio”
    2. Install “R”, 

    Install “RStudio”
    

    View Slide

21. how to install
    R and RStudio
    1. Search “R”, 

    Search “RStudio”
    2. Install “R”, 

    Install “RStudio”
    maybe a few more steps
    so please do this later
    a much better guide:
    rafalab.github.io/dsbook/installing-r-rstudio
    

    View Slide

22. lucky us!
    we have a workaround!
    https://rstudio.cloud
    

    View Slide

23. lucky us!
    we have a workaround!
    https://rstudio.cloud
    do this
    

    View Slide

24. lucky us!
    we have a workaround!
    https://rstudio.cloud this too
    

    View Slide

25. lucky us!
    we have a workaround!
    https://rstudio.cloud
    select
    [from Git repo]
    

    View Slide

26. lucky us!
    we have a workaround!
    https://rstudio.cloud
    enter
    https://github.com/pkimes/dfci-introR
    

    View Slide

27. welcome to
    RStudio!
    

    View Slide

28. local RStudio select
    [New Project]
    

    View Slide

29. local RStudio
    select
    [Version Control]
    

    View Slide

30. local RStudio
    we’ll use [Git]
    

    View Slide

31. local RStudio
    enter
    https://github.com/pkimes/dfci-introR
    

    View Slide

32. we’re good to go!
    

    View Slide

33. you now have a project!
    what’s an RStudio project?
    basically a folder to
    organize an analysis
    • input data
    • R scripts
    • results/figures
    

    View Slide

34. coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    let’s give it a try!
    

    View Slide

35. arithmetic
    variables
    functions
    help
    installing packages
    loading packages
    for-loops
    what did we
    (hopefully) cover?
    

    View Slide

36. some pieces in the modern
    (R) data scientist’s toolbox
    

    View Slide

37. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

38. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

39. .R file
    

    View Slide

40. .R file .Rmd file
    

    View Slide

41. .R file .Rmd file
    formal
    header
    

    View Slide

42. .R file .Rmd file
    code
    “chunks”
    

    View Slide

43. .R file .Rmd file
    plain text
    (markdown)
    

    View Slide

44. .R file .Rmd file
    specified
    output format
    

    View Slide

45. .Rmd file
    

    View Slide

46. .Rmd file
    formatted text
    output!
    R code
    

    View Slide

47. rmarkdown documentation, communication
    myfile.Rmd
    markdown
    +
    R code chunks
    

    View Slide

48. rmarkdown documentation, communication
    myfile.Rmd
    markdown
    +
    R code chunks
    markdown
    myfile.md
    execute
    R code
    

    View Slide

49. rmarkdown documentation, communication
    myfile.Rmd
    markdown
    +
    R code chunks
    markdown
    myfile.md
    execute
    R code
    pandoc
    conversion
    

    View Slide

50. rmarkdown documentation, communication
    rmarkdown.rstudio.com
    

    View Slide

51. coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    coding coding coding coding
    let’s give it a try!
    

    View Slide

52. what did we
    (hopefully) cover?
    create a new Rmd file
    writing simple markdown
    creating code chunks
    executing code
    knitting documents
    

    View Slide

53. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

54. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

55. tidyverse data manipulation, visualization
    tidyverse.org
    

    View Slide

56. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

57. shiny web application framework
    shiny.rstudio.com
    

    View Slide

58. shiny web application framework
    shiny.rstudio.com/gallery/kmeans-example
    

    View Slide

59. some pieces in the modern
    (R) data scientist’s toolbox
    rmarkdown
    tidyverse
    shiny
    [bioconductor]
    documentation, communication
    data manipulation, visualization
    web application framework
    community of genomics packages
    

    View Slide

60. bioconductor community of genomics packages
    bioconductor.org
    

    View Slide

61. bioconductor community of genomics packages
    CRAN Bioconductor
    • genomic focus
    • software
    • annotations
    • data
    • package reviews
    • scope
    • consistency
    

    View Slide

62. awesome!
    

    View Slide

63. where do we go
    from here?
    

    View Slide

64. where do we go
    from here?
    wait, I’m lost
    

    View Slide

65. introduction to data science
    rafalab.github.io/dsbook
    

    View Slide

66. RStudio Cloud tutorials
    rstudio.cloud
    

    View Slide

67. r4ds.had.co.nz
    R for data science
    

    View Slide

68. learn the tidyverse
    tidyverse.org/learn
    

    View Slide

69. advanced R
    adv-r.hadley.nz
    

    View Slide

70. biomedical data science open online training
    rafalab.github.io/pages/harvardx
    

    View Slide

71. … and because it’s 2019, deep learning
    tensorflow.rstudio.com
    

    View Slide

72. questions?
    wait, I’m still lost
    

    View Slide