我正在使用 csv 文件来提取数据并将它们放入字典中以进行分析。CSV 文件位于一个目录中,并且每个目录中都有多个 csv 文件:
Example:
Dir: X has several csv files, file name can be X-ax-somefile.csv, X-bx-somefile2.csv
csv files have the header: level,user-id
我建立了一个字典来存储数据并进行一些计算。最后,我将得到以下数据结构:
{'de': {'en': {'level1': 0, 'level2': 0, 'level3': 10}}, 'en': {'si': {'level2': 1, 'level3': 5, 'level5': 5, 'levelN': 5}, 'en': {'level1': 0}, 'ta': {'level1': 1, 'level2': 1, 'level3': 1, 'level4': 5}}}
我编写了以下代码来迭代这个数据结构,但是,这是最好的迭代方式吗?在这个问题中我已经展示了如何构建数据结构,这是构建数据结构的最佳方式。
这是我的代码:
for lang1, lang2_dict in template_count.iteritems():
if type(lang2_dict):
for lang2, values in lang2_dict.iteritems():
print lang2, values
这就是我构建字典的方式:
def templateUserCountStats(template_file, csv_file):
template_count_dict = dict()
for lang in getLanguageCodes(csv_file):
template_count_dict[lang] = dict()
lang_dir = os.path.join(template_file, lang)
try:
for filename in os.listdir(lang_dir):
path = os.path.join(lang_dir, filename)
if re.search(r'-.+-template-users-data.csv$',filename):
lang2 = filename.split("-")[1]
with open(path, 'rb') as template_user_data_file:
try:
reader = csv.reader(template_user_data_file)
reader.next()
template_count_dict[lang][lang2] = dict()
template_count_dict[lang][lang2]['level1'] = 0
template_count_dict[lang][lang2]['level2'] = 0
template_count_dict[lang][lang2]['level3'] = 0
template_count_dict[lang][lang2]['level4'] = 0
template_count_dict[lang][lang2]['level5'] = 0
template_count_dict[lang][lang2]['levelN'] = 0
print filename
for row in reader:
if int(row[0]) == 1:
template_count_dict[lang][lang2]['level1'] = template_count_dict[lang][lang2]['level1'] + 1
if int(row[0]) == 2:
template_count_dict[lang][lang2]['level2'] = template_count_dict[lang][lang2]['level2'] + 1
if int(row[0]) == 3:
template_count_dict[lang][lang2]['level3'] = template_count_dict[lang][lang2]['level3'] + 1
if int(row[0]) == 4:
template_count_dict[lang][lang2]['level4'] = template_count_dict[lang][lang2]['level4'] + 1
if int(row[0]) == 5:
template_count_dict[lang][lang2]['level5'] = template_count_dict[lang][lang2]['level5'] + 1
if row[0] == 'N':
template_count_dict[lang][lang2]['levelN'] = template_count_dict[lang][lang2]['levelN'] + 1
except csv.ERROR as e:
logging.error(e)
else:
continue
except Exception, e:
logging.exception(e)
return template_count_dict
除了我愿意接受有关构建此数据结构的建议之外,这是一种更加 Python 的方式,如果您需要以下示例:
level,user-id
1,25
1,74
1,105
3,708
3,530
3,2568
3,2730
3,2730
2,376
2,371
2,2317
2,2095
N,560
N,110
N,119
N,1059
N,1625