backward elimination in R

When comparing different submodels, it is necessary that they be fitted to the same set of data — otherwise the results just don’t make sense. (Consider the extreme situation where you have two predictors A and B, which are each measured on only half of your observations — then the model y~A+B will be fitted to all the data, but the models y~A and y~B will be fitted to non-overlapping subsets of the data.) Thus, step won’t allow you to compare submodels that (because of automatic removal of cases containing NA values) are using different subsets of the original data set.

Using na.omit on the original data set should fix the problem.

fullmodel <- lm(Eeff ~ NDF + ADF + CP + NEL + DMI + FCM, data = na.omit(phuong))
step(fullmodel, direction = "backward", trace=FALSE )

However, if you have a lot of NA values in different predictors, you may end up losing a lot of your data set — in an extreme case you could lose the entire data set. If this happens you have to reconsider your modeling strategy …

Related Posts:

  1. Plot multiple lines in one graph
  2. Create empty data frame with column names by assigning a string vector?
  3. Changing font size and direction of axes text in ggplot2
  4. How to randomize ggplot factor color scheme, keeping same default distances on color wheel?
  5. Warning message: line appears to contain embedded nulls [duplicate]
  6. Find duplicate values in R
  7. Error in library(ggplot2) : There is no package called ‘ggplot2’
  8. Transpose a data frame
  9. How to drop columns by name in a data frame
  10. How to reshape data from long to wide format
  11. Change size of axes title and labels in ggplot2
  12. Multiple If Statements in R
  13. ggplot geom_text font size control
  14. What’s the difference between lapply and do.call?
  15. ggplot2, change title size
  16. Initial value in ‘vmmin’ is not finite even when changing the starting value
  17. Insert picture/table in R Markdown
  18. How to remove outliers from a dataset
  19. Mean per group in a data.frame
  20. Error in dev.off() : cannot shut down device 1 (the null device)
  21. How to create an empty matrix in R?
  22. standard deviation on dataframe does not work
  23. How to increase font size in a plot in R?
  24. What is the difference between = and ==?
  25. Merge r brings error “‘by’ must specify uniquely valid columns”
  26. Subscripts in plots in R
  27. promise already under evaluation: recursive default argument reference or earlier problems?
  28. Aggregate multiple columns at once
  29. Error in model.frame.default: variable lengths differ
  30. R – Concatenate two dataframes?
  31. KNN in R: ‘train and class have different lengths’?
  32. Producing subscripts in R markdown
  33. ggplot geom_text font size control
  34. How do you use “<<-" (scoping assignment) in R?
  35. Re-ordering factor levels in data frame
  36. Is it possible to have a multi-line comments in R?
  37. Understanding the result of modulo operator: %%
  38. R – error “variable lengths differ”
  39. Simple manual RMarkdown tables that look good in HTML, PDF and DOCX
  40. Select rows from a data frame based on values in a vector
  41. how to increase the limit for max.print in R
  42. dplyr: “Error in n(): function should not be called directly”
  43. In R formulas, why do I have to use the I() function on power terms, like y ~ I(x^3)
  44. Filter data.frame rows by a logical condition
  45. How do I convert a factor into date format?
  46. Error while creating heatmaps – NA/NaN/Inf in foreign function call (arg 11)
  47. Write lines of text to a file in R
  48. How to compute summation in r
  49. Plot multiple boxplot in one graph
  50. R – argument is of length zero in if statement
  51. ‘Incomplete final line’ warning when trying to read a .csv file into R
  52. How to solve ‘protection stack overflow’ issue in R Studio
  53. Conditional mean statement
  54. Merging data – Error in fix.by(by.x, x)
  55. Warning in GLM analysis
  56. Error in 2:n : NA/NaN argument
  57. Error in generating boxplot two variable data frame: adding class “factor” to an invalid object
  58. What does ‘length of ‘dimnames’ [1] not equal to array extent’ mean?
  59. How do I flip rows and columns in R
  60. How to catch integer(0)?
  61. Print string and variable contents on the same line in R
  62. “no function to return from, jumping to top level”
  63. Does R have a wildcard expression (such as an asterisk (*))?
  64. r random forest error – type of predictors in new data do not match
  65. env=baseenv() in R
  66. How to set axis range R
  67. Simple for loop in R producing “replacement has length zero” in R
  68. How to find out which package version is loaded in R?
  69. How to solve it. error:Error in data[[rowvar]] : attempt to select less than one element
  70. What does Continuous x aesthetic — did you forget aes(group=…) mean?
  71. How to unload a package without restarting R
  72. What does the error “object not interpretable as a factor” mean?
  73. use first row data as column names in r
  74. Gradient of n colors ranging from color 1 and color 2
  75. How to prevent scientific notation in R?
  76. What is the difference between require() and library()?
  77. Launching R help: Error in file(out, “wt”) : cannot open the connection
  78. R Sweave: NO TeX installation detected
  79. Plotting legend outside plot in R
  80. Reset par to the default values at startup
  81. How to increase the font size of ggtitle in ggplot2
  82. ERROR: [on_request_read] connection reset by peer in R shiny
  83. Error in G * t(hat) : non-conformable arrays
  84. Get dplyr count of distinct in a readable way
  85. Error: isTRUE(gpclibPermitStatus()) is not TRUE
  86. How to counter the ‘non-numeric matrix extent’ error in R?
  87. Merge two rows in data.frame
  88. r function unzip error 1 in extracting from zip file
  89. Get the list of installed packages by user in R
  90. Colour points in a plot differently depending on a vector of values
  91. How to access the last value in a vector?
  92. How To Create Vector of Vector In R
  93. Use of ~ (tilde) in R programming Language
  94. Error: x must be atomic for ‘sort.list’
  95. Transposing in dplyr
  96. Does the c command create a row vector or a column vector by default in R
  97. Can dplyr join on multiple columns or composite key?
  98. Error in summary.connection(connection) : invalid connection
  99. Converting from a character to a numeric data frame
  100. How to create lag variables

Leave a Comment