====== Listing a directory contents ======
In your solution, you will have to go through all files stored in directories:
  * images in training data directory (to learn something from them, i.e., estimate the parameters of a model) and
  * images in testing data directory (to classify these images with the chosen classifier).
There is more ways to achieve this; we shall list here some of them.

===== os.listdir() =====
Function ''listdir()'' ([[https://docs.python.org/3/library/os.html?highlight=listdir#os.listdir|docs]]) from ''os'' module returns a list of filename from specified directory. If we have the following directory with files
<code>
+- train_data
   +- truth.dsv
   +- img112.png
   +- img113.png
   +- img114.png
</code>
then the following script
<code python>
import os

for fname in os.listdir("train_data"):
    print(fname)
</code>
provides this output:
<code>
img_1112.png
img_1113.png
img_1114.png
truth.dsv
</code>

You can also use similar function [[https://docs.python.org/3/library/os.html?highlight=listdir#os.scandir|os.scandir()]] which returns a generator.

===== Object-oriented interface of ''pathlib'' module =====
If you like object-oriented interfaces better, then you can use class ''Path'' and its method ''iterdir()'' ([[https://docs.python.org/3/library/pathlib.html?highlight=iterdir#pathlib.Path.iterdir|docs]]):
<code python>
from pathlib import Path

path = Path("train_data")
for fpath in path.iterdir():
    print(fpath)
</code>

After running the above script:

<code>
train_data\img_1112.png
train_data\img_1113.png
train_data\img_1114.png
train_data\truth.dsv
</code>

Generator ''iterdir()'' returns instances of ''Path'' class which encapsulate the paths to individual files. But you can easily use these objects when opening files using ''open()'' function.

More examples how to use ''pathlib'' can be found in this nice [[https://miguendes.me/python-pathlib|pathlib tutorial]].