Python function os.listdir 不显示目录中的所有文件

Question

Hi I'm trying load couple of files with os.listdir but it skips several files.

Format of files in directiy is:

a_0.csv
a_1.csv
a_2.csv
  ·
  ·
  ·
a_25.csv

But the os.listdir only take file 0,1,10,11.. File with numbers 3,4,5... aren't in in output object.

And here is my function:

def best_fit_from_attemp(self):
    arr =[]
    for i in os.listdir(self.path_to_files):
        print("i",i)
        if i[0] == 'f' and  i[1] == 'i' and i[2] == 't':
            print("searching for smallest number... ",i)
            arr.append(self.find_min_fitness(i))

Output:

i fitness_gen_0.csv
searching for smallest number...  fitness_gen_0.csv
i fitness_gen_1.csv
searching for smallest number...  fitness_gen_1.csv
i fitness_gen_10.csv
searching for smallest number...  fitness_gen_10.csv
i fitness_gen_11.csv
searching for smallest number...  fitness_gen_11.csv

Answer 1

My guess is that the filenames are all there, just not in the order you expect.

The file names are strings. They are sorted according to string-sorting logic: imagine replacing every 0 with an a , every 1 with a b , every 2 with a c and so on, and now think about how the strings would be ordered alphabetically. Ordinary string sorting doesn't know that we like to write numbers with the least-significant-digit last . In ordinary string sorting, it really does go 1 , then 10 , then 11 —that's analogous to ordering the strings b , then ba , then bb , which is the correct dictionary ordering. It will go on for a long time before it gets to anything starting with 3 (analogously, the words in the dictionary that start with d come much later than anything starting with b ).

You could investigate algorithms/packages that do "natural sorting". Or, the greatly preferable best practice would be simply to fix your file naming convention to start with: ensure that the numbers embedded in your filenames all have the same number of digits, with leading zeros where necessary ( 001 , 002 , etc).

Answer 2

I would be pretty sure that the problem lies in the function self.find_min_fitness(i) . Some parts in it must cause the program to break after listing the first couple of files.

The files are sorted by name by os.listdir() . It prefers 1 over 2 when listing the files, that's why all the filenames starting with 1 are listed first.

As a solution: Maybe try and comment out this one line where self.find_min_fitness(i) is called and see whether the problem still occurs

Answer 3

This seems questionable.

self.find_min_fitness(i)

Well I would suggest you try something like this.

Below directory contains a list of files numbered.

files = os.listdir("C:/Users/Danish/Documents/JavaScriptPrimerRepo/MyJSCodes")

for i in files:
    print(i)

This is the sample output below.

00_BlankPage.html
01_DataType.html
02_LinkingExternalJsfile.html
02_Script.js
03_VariableMutation_cohersion.html
04_Operators.html
05_SampleCodeCalculatingBMI.html

NOTE: I stored the result in separate variable before iterating and importantly fileName starts with number first.