-
Notifications
You must be signed in to change notification settings - Fork 5
/
banking.py
123 lines (107 loc) · 4.23 KB
/
banking.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
import csv
import requests
import urllib.request
import zipfile
import os
from urllib.request import urlretrieve
from zipfile import ZipFile
import pandas as pd
district_column_names = [
"district_id",
"district_name",
"region",
"no_of_inhabitants",
"no_of_municipalities_lt_499",
"no_of_municipalities_500_1999",
"no_of_municipalities_2000_9999",
"no_of_municipalities_gt_10000",
"no_of_cities",
"ratio_of_urban_inhabitants",
"average_salary",
"unemployment_rate_95",
"unemployment_rate_96",
"no_of_entrepreneurs_per_1000_inhabitants",
"no_of_committed_crimes_95",
"no_of_committed_crimes_96",
]
class BankingData:
def __init__(self, url, data_name):
self.url = url
self.data_name = data_name
def extract_to_csv(self):
# check if "bank_data" folder exists, if not, create it
if not os.path.exists("bank_data"):
os.mkdir("bank_data")
# Retrieve the zip file from the url link
file = os.path.basename(self.url)
urlretrieve(self.url, file)
# Extract the zip file's contents
with ZipFile(file, "r") as zf:
zf.extractall("bank_data")
# The file containing our data
csv_file_name = f"{self.data_name}.csv"
# Data clean up
df = pd.read_csv(f"bank_data/{csv_file_name}", sep=";")
# Save the cleaned up CSV file
df.to_csv(df.to_csv(f"{self.data_name}_cleaned.csv", index=False))
class MarketData:
def __init__(self, url, output_folder):
self.url = url
self.output_folder = output_folder
def extract(self):
"""
This function extracts the banking data provided from PKDD.
It downloads the ZIP file from the "url".
Args:
url (str): the URL containing the public data
"""
try:
response = requests.get(self.url, stream=True)
zip_file_path, _ = urllib.request.urlretrieve(self.url)
with open(zip_file_path, "wb") as out_file:
for chunk in response.iter_content(chunk_size=8192):
out_file.write(chunk)
# Now, try to extract the ZIP file
with zipfile.ZipFile(zip_file_path, "r") as zip_ref:
zip_ref.extractall(self.output_folder)
return zip_ref
except Exception as e:
print(f"Error, could not download data: {e}")
def convert_asc_to_csv(self, district_column_names):
"""
This function converts the .asc files to the .csv format.
The function outputs a folder with a name from output_folder.
This created folder will be in the current directory.
Args:
output_folder (str): the name of the folder where
files will be stored
"""
try:
zip_ref = self.extract()
# Process ASC files and convert them to CSV
for file_name in zip_ref.namelist():
if file_name.endswith(".asc"):
asc_path = os.path.join(self.output_folder, file_name)
csv_file_name = file_name[:-4] + ".csv"
csv_path = os.path.join(self.output_folder, csv_file_name)
with open(asc_path, "r") as asc_file, open(
csv_path, "w", newline=""
) as csv_file:
asc_reader = csv.reader(asc_file, delimiter=";")
csv_writer = csv.writer(csv_file, delimiter=",")
if file_name == "district.asc":
next(asc_reader)
new_header = district_column_names
csv_writer.writerow(new_header)
csv_writer.writerows(asc_reader)
else:
for row in asc_reader:
csv_writer.writerow(row)
print(f"Converted {asc_path} to CSV.")
print("All ASC files converted to CSV.")
except Exception as e:
print(f"Error, could not convert ASC to CSV: {e}")
# Example usage
# link = "http://sorry.vse.cz/~berka/challenge/pkdd1999/data_berka.zip"
# output = "expanded_data"
# extract_asc_to_csv(link, output)