The Process
I think I enjoyed this week's exercises quite a lot. I have used a few before, such as ls, pwd, and a few other
basic ones, but this really opened up to me how powerful command line is, and how the shell actually works and processes
these commands. The challenges were fairly simple, but let me really practice how to use the command line.
I also googled around online a lot, and found plenty of really helpful resources!
Challenges
1. sort vs sort -n
This challenge was one which we mentioned briefly in class. By default, the sort command is using the first integer
it runs across as the number upon which it sorts. Meaning for "10", "100", and "1000", it is only taking into account the
"1" when sorting. So when you have any other number, such as "2", the sort command is only comparing the "2" and the "1"
digit and finds that the "2" is higher. On the other hand, sort -n tells the program to sort by arithmetic value, and
so will take into account the full number, as well as minus signs and decimal digits.
2. wc -l < mydata.dat vs wc -l mydata.dat
In wc -l < mydata.dat , wc is not given any command line parameters, and so defaults to reading from standard input. But,
we have told shell to send the contents of mydata.dat to the standard input which wc will use.
In wc -l mydata.dat, wc is given a command line parameter telling it to open data from mydata.dat.
I have found that both will output the number of lines within a file, but wc -l mydata.dat will list the mydata.dat
filename which was specified as the parameter, while wc -l < mydata.dat will not specify the mydata.dat which was sent to the
standard input.
3. Uniq
Now this might be wrong, but it seems to me that the uniq command would be useful when cleaning up data sets and only
wanting to see every instance when there is a change, and not necessarily every single instance of something occurring. I've
done some psychology research, and what we were concerned with was how often a participant switched two presented rewards,
and not necessarily how many times they chose a certain reward consecutively. If data had been collected for every single choice,
uniq would have been a great command to clean up the data. I'm guessing uniq could also get rid of accidental double entries
in large data sets.
If you wanted to remove ALL duplicated lines, then you could use sort first in order to have all the data sets sorted, with duplicates
listed adjacently, and then | uniq to remove all of the duplicates, which would be located adjacent.
4. animals.txt
Time to break this up into pieces. The command is cat animals.txt | head -5 | tail -3 | sort -r > final.txt .
In the first part cat animals.txt, cat is concatenating the contents of the files defined in the parameter. Since animals.txt
is the only defined file, it just lists the contents of animals.txt. So the output for this portion should be list:
2012-11-05,deer
2012-11-05,rabbit
2012-11-05,raccoon
2012-11-06,rabbit
2012-11-06,deer
2012-11-06,fox
2012-11-07,rabbit
2012-11-07,bear
head -5 means we want the first 5 lines of the file, meaning the output is list:
2012-11-05,deer
2012-11-05,rabbit
2012-11-05,raccoon
2012-11-06,rabbit
2012-11-06,deer
tail -3 displays the last 3 lines of a file, meaning the output is list:
2012-11-05,raccoon
2012-11-06,rabbit
2012-11-06,deer
sort -r > final.txt is doing 2 things. First, sort -r is reversing the order of the specified sort. > final.txt
is a command to find a folder final.txt and overwrite it with this new data, or create and save this new data if
final.txt did not originally exist. The data saved in file final.txt should be list:
2012-11-06,rabbit
2012-11-06,deer
2012-11-05,raccoon
5. What animals does the file contain? (No duplicates)
Given the command cut -d , -f 2 animals.txt, we can use the uniq and sort command to remove all duplicates.
cut -d , -f 2 animals.txt | sort | uniq, which would return a list"
bear
deer
fox
rabbit
raccoon