I know your job is implementing Machine Learning(ML) algorithms and you may think, learning Linux command is irrelevant but based on my experience you should know all ins-and-outs of an ML project, including understanding, organizing and preparing data before feeding the data into your ML network. So, I am writting several posts about Linux commands that can be so practical and helpful for data preparation as a part of doing an ML project.

How to see the list of directories under Linux operating systems?

You need to use a Linux command called "tree". It will list contents of directories in a tree-like format. It is a recursive directory listing program that produces a depth indented listing of files. When directory arguments are given, tree lists all the files and/or directories found in the given directories each in turn. tree returns the total number of files and/or directories listed.

How to install "tree" on a Linux?

# yum install tree

The syntax is:

tree

tree /path/to/directory

tree [options]

tree [options] /path/to/directory

Here is a list of all options with a short description supported by the tree program:

  ------- Listing options -------
  -a            All files are listed.(except hidden files)   -d            List directories only.
  -l            Follow symbolic links like directories.
  -f            Print the full path prefix for each file.
  -l            Follow symbolic links like directories.
  -L level      Descend only level directories deep.
  -R            Rerun tree when max dir level reached.
  -P pattern    List only those files that match the pattern given.
  -I pattern    Do not list files that match the given pattern.
  --matchdirs   Include directory names in -P pattern matching.      --noreport    Turn off file/directory count at end of tree listing.
  --charset X   Use charset X for terminal/HTML and indentation line output.
  --filelimit # Do not descend dirs with more than # files in them.
  --timefmt <f> Print and format time according to the format <f>.
  -o filename   Output to file instead of stdout.

  -------- File options ---------
  -q            Print non-printable characters as '?'.
  -N            Print non-printable characters as is.
  -Q            Quote filenames with double quotes.
  -p            Print the protections for each file.
  -u            Displays file owner or UID number.
  -g            Displays file group owner or GID number.
  -s            Print the size in bytes of each file.
  -h            Print the size in a more human readable way.
  --si          Like -h, but use in SI units (powers of 1000).
  -D            Print the date of last modification or (-c) status change.
  -F            Appends '/', '=', '*', '@', '|' or '>' as per ls -F.
  --inodes      Print inode number of each file.
  --device      Print device ID number to which each file belongs.

  ------- Sorting options -------
  -v            Sort files alphanumerically by version.
  -t            Sort files by last modification time.
  -c            Sort files by last status change time.
  -U            Leave files unsorted.
  -r            Reverse the order of the sort.
  --dirsfirst   List directories before files (-U disables).
  --sort X      Select sort: name,version,size,mtime,ctime.

The most intersting part of tree command for me is pattern option, "-P". Since I just need ".jpg" files to feed my network, I can use tree command with -p parameter to show me the right directories containing ".jpeg" files.

Here is a simple example of using pattern option on a Linux tree command:

$ sudo tree -f -P *.jpg

So, you can find the directories contain images you want to download from them.

I hope you find this post helpful as one of the beginning steps for the data preparation process.

# Plant Classification

LLinux tree command

HHybrid approches using for plant classification