Due: Wednesday, February 12, 8pm
As we've discussed, for each homework please create a new,
clearly named subfolder in your Google Drive folder (for
example, Unit 1, Lesson 3
) and put any files for the assignment in there. Please
create a new Python file
(.py) for each homework
part and name the files accordingly. For
example unit1_hw3_part1.py,
or if your folder is accurately named, simply name the
file part1.py.
Ambiguously labeled homework folders or files may not
receive credit if I cannot identify which assignment they
are meant to fulfill. Thank you.
Work with user input as data for an
algorithm. Create a new file and add the Python code
from Unit 1,
Lesson 2 that finds the smallest number in a list, then
referring back to your work for the Lesson 2 Homework, modify
this so it finds the alphabetically last word in a list. Now,
at the top of the file, add import sys, and then
use sys.argv as the data list to run your
algorithm on. (See the
notes
on command line arguments for help.) Then, run your
program from the command line specifying 5 (or more)
words. Run the program at least two times to see the different
output. Include a screenshot of the output in your homework
folder this week, clearly labeled as "Part 1 output".
One problem you might see is that the alphabetically last word
might be the name of your Python file itself. Remember that
the first item in the sys.argv list is the name
of the program you are running. To change this so your program
only considers the user inputs, you can remove the first item
in the list. The command to do that
is pop(0). (You can read
the Python
documentation on pop() here.) So after you
import sys, you might have the following two
lines:
word_list = sys.argv word_list.pop(0)
Allow the user to do some basic image manipulation from the command line. Create a new file for this part. Refer back to the section of the class notes that describes how to take a command line argument and open an image file. Use the version of this code the includes some error handling.
The Python Image Library
command rotate()
allows you to rotate an image within Python by specifying a
number which is degrees. Modify your error handling so it
requires two command line arguments
(i.e. len(sys.argv) should be 3). Treat the first
as a filename and open an image file, as we have done. Then
treat the second argument as a number to pass in
to rotate(). You will need to convert the
argument, which is always text, into a number with
the int() command. The following line should do
it:
rotated_img = img.rotate( int(sys.argv[2]) )
The rotate() command returns a new
image, which is why we have to create a new variable here
which I'm calling rotated_img and assign the
result of rotate() to that new variable.
The last thing we'll need to do is save this rotated
image. The PIL command save() does this for
us. (Read more
documentation here.)
Save the new image we created, rotated_img, like
this:
rotated_img.save("rotated-" + sys.argv[1])
Now try running this a couple times with different filenames and numerical values for angles to rotate by. Include a sample output in your homework folder.
Create a new file. To start, again refer back to the section of the class notes that describes how to take a command line argument and open an image file and use the version of this code the includes some error handling. Modify the error handling so that your code requires two filenames.
Open the first filename as an image as before. Copy/paste that line so your code also opens the second filename as an image.
Now let's use the Python Image Library
command blend(). (See
the blend()
documentation for reference.) The command takes two images
and a number between 0 and 1 which corresponds to how much
transparency to use in blending them. As
with rotate(), this returns a new image, so we'll
have to make a new variable to hold it. The line should look
something like this:
blended_img = Image.blend(img1,img2,.5)
For that line to work you will have had to name your two
image variables img1 and img2.
Now save that new image variable. (See the save()
line in part two above, but for the filename,
use "blended.jpg" and you don't need to
reference sys.argv at all.)
Try running this a few times. You will need to reference images in your current directory, so if you'd like to experiment, perhaps add some additional images there. Include some sample outputs in your homework folder.
[Note added after
class.] As we discussed in class, many of you
ran into some issues here, encountering errors
like: ValueError: images do not match. The
solution here is to make sure that your images are using the
same internal color mode, and possibly also
to make sure that the images are the same size. You can read
more about color modes in the Python Image
Library documentation
here: Concepts:
Modes. To change the color mode of your
image, use the
PIL convert()
command. Note that, similar to blend()
and rotate(), the convert()
command returns a new image —
it does not modify your existing
image in-place. That means you have to take
the result of the command and assign it to a new
variable. Like so:
converted_img = img.convert("RGB")
Repeat that command for both images that you are working with,
and then try to blend() them together.
If you are still getting a ValueError, you may
need to also resize the images so they are the same size. You
can do this with
the resize()
command, which, again returns a new image, rather than
modifying the image in-place. This command
does that funny PIL thing where it requires one argument which
is a pair of values, so they must be enclosed in parentheses. Like this:
resized_img = img.resize( (400,400) )The
(400,400) was arbitrary. Experiment with
different values as you'd like. The first number
corresponds to the width of the image, and the second to the
height. The only caveat is to make sure that you are using the
same size for both images.
You can combine convert()
and resize() in one of two ways. One separate
lines, like this:
converted_img = img.convert("RGB")
resized_img = converted_img.resize( (400,400) )
or on one line, like this:
resized_img = img.convert("RGB").resize( (400,400) )
Note that in the first example, I'm using the output of
the first command, which is saved into the variable I
named converted_img, as the variable on which to
call the resize() command.
In the second example above, I'm
calling resize() on the output
of convert() directly, by chaining commands
together with the dot
notation ..
(Challenge.) Can you modify the above code so it only
takes one argument, which corresponds to the name of a
subfolder containing images. Then randomly select two files
from that folder and blend() them
together. This will be key for doing the Unit 1 project, so
we'll talk about it more in the coming weeks. Consider
trying this now as a challenge. For help on how to randomly
select images from a subfolder, take a look at this Gist:
Random file selection help