Create dataframe from a matrix

Question

How to get a data frame with the same data as an already existing matrix has   A simplified example of my matrix   mat  lt - matrix c 0  0 5  1  0 1  0 2  0 3  0 3  0 4  0 5                 ncol   3  nrow   3                dimnames   list NULL  c  time    C 0    C 1       gt  mat      time C 0 C 1  1    0 0 0 1 0 3  2    0 5 0 2 0 4  3    1 0 0 3 0 5   I would like to create a data frame that looks like this        name   time   val 1    C 0    0 0    0 1 2    C 0    0 5    0 2 3    C 0    1 0    0 3 4    C 1    0 0    0 3 5    C 1    0 5    0 4 6    C 1    1 0    0 5   All my attempts are quite clumsy  for example   data frame cbind c rep  C 1   3   rep  C 2   3                     rbind cbind mat   time    mat   C 0                            cbind mat   time    mat   C 1         Does anyone have an idea of how to do this more elegantly  Please note that my real data has a few more columns  40 columns

User · Answer

I ve found the following  cheat  to work very neatly and error-free   gt  dimnames  lt - list time c 0  0 5  1   name c  C 0    C 1     gt  mat  lt - matrix data  ncol 2  nrow 3  dimnames dimnames   gt  head mat  2   this returns the number of rows indicated in a data frame format  gt  df  lt - data frame head mat  2     data frame  might not be necessary   Et voila

User · Answer

Using dplyr and tidyr   library dplyr  library tidyr   df  lt - as data frame mat    gt          convert the matrix to a data frame   gather name  val  C 0 C 1    gt      convert the data frame from wide to long   select name  time  val            reorder the columns  df   A tibble  6 x 3    name  time   val    lt chr gt   lt dbl gt   lt dbl gt  1   C 0   0 0   0 1 2   C 0   0 5   0 2 3   C 0   1 0   0 3 4   C 1   0 0   0 3 5   C 1   0 5   0 4 6   C 1   1 0   0 5

User · Answer

If you change your time column into row names  then you can use as data frame as table mat   for simple cases like this   Example   data  lt - c 0 1  0 2  0 3  0 3  0 4  0 5  dimnames  lt - list time c 0  0 5  1   name c  C 0    C 1    mat  lt - matrix data  ncol 2  nrow 3  dimnames dimnames  as data frame as table mat     time name Freq 1    0  C 0  0 1 2  0 5  C 0  0 2 3    1  C 0  0 3 4    0  C 1  0 3 5  0 5  C 1  0 4 6    1  C 1  0 5   In this case time and name are both factors  You may want to convert time back to numeric  or it may not matter

User · Answer

You can use stack from the base package  But  you need first to coerce your matrix to a data frame and to reorder the columns once the data is stacked   mat  lt - as data frame mat  res  lt - data frame time  mat time stack mat select -time   res  c 3 1 2      ind time values 1 C 0  0 0    0 1 2 C 0  0 5    0 2 3 C 0  1 0    0 3 4 C 1  0 0    0 3 5 C 1  0 5    0 4 6 C 1  1 0    0 5   Note that stack is generally more efficient than the reshape2 package

User · Answer

melt   from the reshape2 package gets you close       library reshape2   res  lt - melt as data frame mat   id  time        time variable value   1  0 0      C 0   0 1   2  0 5      C 0   0 2   3  1 0      C 0   0 3   4  0 0      C 1   0 3   5  0 5      C 1   0 4   6  1 0      C 1   0 5       although you may want to post-process its results to get your preferred column names and ordering   setNames res c  variable    time    value     c  name    time    val        name time val   1  C 0  0 0 0 1   2  C 0  0 5 0 2   3  C 0  1 0 0 3   4  C 1  0 0 0 3   5  C 1  0 5 0 4   6  C 1  1 0 0 5

[r] Create dataframe from a matrix

Examples related to r

Examples related to matrix

Examples related to dataframe