You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
403 lines
14 KiB
Plaintext
403 lines
14 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 143,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import pandas as pd\n",
|
|
"import pyarrow\n",
|
|
"from tqdm import tqdm\n",
|
|
"import os\n",
|
|
"import shutil"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": "'2.0.0rc1'"
|
|
},
|
|
"execution_count": 2,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"pd.__version__"
|
|
],
|
|
"metadata": {
|
|
"collapsed": false
|
|
}
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 12,
|
|
"outputs": [
|
|
{
|
|
"ename": "OptionError",
|
|
"evalue": "'You can only set the value of existing options'",
|
|
"output_type": "error",
|
|
"traceback": [
|
|
"\u001B[1;31m---------------------------------------------------------------------------\u001B[0m",
|
|
"\u001B[1;31mOptionError\u001B[0m Traceback (most recent call last)",
|
|
"Cell \u001B[1;32mIn[12], line 1\u001B[0m\n\u001B[1;32m----> 1\u001B[0m \u001B[43mpd\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43moptions\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mmode\u001B[49m\u001B[38;5;241;43m.\u001B[39;49m\u001B[43mdtype_backend\u001B[49m \u001B[38;5;241m=\u001B[39m \u001B[38;5;124m'\u001B[39m\u001B[38;5;124mpyarrow\u001B[39m\u001B[38;5;124m'\u001B[39m\n",
|
|
"File \u001B[1;32m~\\.conda\\envs\\MOME_BIGDATA\\lib\\site-packages\\pandas\\_config\\config.py:226\u001B[0m, in \u001B[0;36mDictWrapper.__setattr__\u001B[1;34m(self, key, val)\u001B[0m\n\u001B[0;32m 224\u001B[0m _set_option(prefix, val)\n\u001B[0;32m 225\u001B[0m \u001B[38;5;28;01melse\u001B[39;00m:\n\u001B[1;32m--> 226\u001B[0m \u001B[38;5;28;01mraise\u001B[39;00m OptionError(\u001B[38;5;124m\"\u001B[39m\u001B[38;5;124mYou can only set the value of existing options\u001B[39m\u001B[38;5;124m\"\u001B[39m)\n",
|
|
"\u001B[1;31mOptionError\u001B[0m: 'You can only set the value of existing options'"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# pd.options.mode #= 'pyarrow'"
|
|
],
|
|
"metadata": {
|
|
"collapsed": false
|
|
}
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 119,
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_01\n",
|
|
"D:\\PATSTAT\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"import os\n",
|
|
"print(os.getcwd()) # Prints the current working directory\n",
|
|
"\n",
|
|
"workdir_path=r\"D:\\PATSTAT\"\n",
|
|
"os.chdir(workdir_path)\n",
|
|
"print(os.getcwd())"
|
|
],
|
|
"metadata": {
|
|
"collapsed": false
|
|
}
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"outputs": [],
|
|
"source": [],
|
|
"metadata": {
|
|
"collapsed": false
|
|
}
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 148,
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"CONCAT: ['tls201_part01.csv', 'tls201_part02.csv', 'tls201_part03.csv'] TO D:\\PATSTAT\\table_tls201.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"1000it [07:12, 2.31it/s]\n",
|
|
"1000it [07:25, 2.25it/s]\n",
|
|
"429it [02:55, 2.44it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_01\\tls201_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_01\\tls201_part02.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_01\\tls201_part03.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_01\\tls202_part01.csv TO D:\\PATSTAT\\table_tls202.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_05\\tls204_part01\\tls204_part01.csv TO D:\\PATSTAT\\table_tls204.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_05\\tls205_part01\\tls205_part01.csv TO D:\\PATSTAT\\table_tls205.csv\n",
|
|
"CONCAT: ['tls206_part01.csv', 'tls206_part02.csv'] TO D:\\PATSTAT\\table_tls206.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"1000it [07:21, 2.27it/s]\n",
|
|
"715it [05:32, 2.15it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_05\\tls206_part01\\tls206_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_05\\tls206_part02\\tls206_part02.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_05\\tls207_part01\\tls207_part01.csv TO D:\\PATSTAT\\table_tls207.csv\n",
|
|
"CONCAT: ['tls209_part01.csv', 'tls209_part02.csv'] TO D:\\PATSTAT\\table_tls209.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"3200it [07:35, 7.02it/s]\n",
|
|
"3049it [07:15, 7.01it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls209_part01\\tls209_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls209_part02\\tls209_part02.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls210_part01\\tls210_part01.csv TO D:\\PATSTAT\\table_tls210.csv\n",
|
|
"CONCAT: ['tls211_part01.csv', 'tls211_part02.csv'] TO D:\\PATSTAT\\table_tls211.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"2500it [08:45, 4.76it/s]\n",
|
|
"369it [01:23, 4.42it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls211_part01\\tls211_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls211_part02\\tls211_part02.csv\n",
|
|
"CONCAT: ['tls212_part01.csv', 'tls212_part02.csv', 'tls212_part03.csv'] TO D:\\PATSTAT\\table_tls212.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"4000it [12:45, 5.23it/s]\n",
|
|
"4000it [12:57, 5.14it/s]\n",
|
|
"1232it [03:54, 5.25it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_06\\tls212_part01\\tls212_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_07\\tls212_part02\\tls212_part02.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_07\\tls212_part03\\tls212_part03.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_07\\tls214_part01\\tls214_part01.csv TO D:\\PATSTAT\\table_tls214.csv\n",
|
|
"CONCAT: ['tls215_part01.csv', 'tls215_part02.csv', 'tls215_part03.csv'] TO D:\\PATSTAT\\table_tls215.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"8000it [15:10, 8.79it/s]\n",
|
|
"8000it [14:20, 9.30it/s]\n",
|
|
"1294it [02:10, 9.95it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls215_part01\\tls215_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls215_part02\\tls215_part02.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls215_part03\\tls215_part03.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls216_part01\\tls216_part01.csv TO D:\\PATSTAT\\table_tls216.csv\n",
|
|
"CONCAT: ['tls222_part01.csv', 'tls222_part02.csv'] TO D:\\PATSTAT\\table_tls222.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"5000it [09:11, 9.06it/s]\n",
|
|
"2402it [04:24, 9.07it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls222_part01\\tls222_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls222_part02\\tls222_part02.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls223_part01\\tls223_part01.csv TO D:\\PATSTAT\\table_tls223.csv\n",
|
|
"CONCAT: ['tls224_part01.csv', 'tls224_part02.csv', 'tls224_part03.csv'] TO D:\\PATSTAT\\table_tls224.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"3000it [04:43, 10.60it/s]\n",
|
|
"3000it [04:44, 10.53it/s]\n",
|
|
"1131it [01:48, 10.38it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls224_part01\\tls224_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls224_part02\\tls224_part02.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls224_part03\\tls224_part03.csv\n",
|
|
"CONCAT: ['tls225_part01.csv', 'tls225_part02.csv'] TO D:\\PATSTAT\\table_tls225.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"3000it [09:19, 5.37it/s]\n",
|
|
"529it [01:39, 5.32it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls225_part01\\tls225_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_08\\tls225_part02\\tls225_part02.csv\n",
|
|
"CONCAT: ['tls226_part01.csv', 'tls226_part02.csv'] TO D:\\PATSTAT\\table_tls226.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"1500it [11:54, 2.10it/s]\n",
|
|
"581it [04:50, 2.00it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_09\\tls226_part01\\tls226_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_09\\tls226_part02\\tls226_part02.csv\n",
|
|
"CONCAT: ['tls227_part01.csv', 'tls227_part02.csv'] TO D:\\PATSTAT\\table_tls227.csv\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"8000it [11:59, 11.12it/s]\n",
|
|
"434it [00:39, 11.09it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"deleting\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_09\\tls227_part01\\tls227_part01.csv\n",
|
|
"D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_09\\tls227_part02\\tls227_part02.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_09\\tls228_part01\\tls228_part01.csv TO D:\\PATSTAT\\table_tls228.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls229_part01\\tls229_part01.csv TO D:\\PATSTAT\\table_tls229.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls230_part01\\tls230_part01.csv TO D:\\PATSTAT\\table_tls230.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls801_part01\\tls801_part01.csv TO D:\\PATSTAT\\table_tls801.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls803_part01\\tls803_part01.csv TO D:\\PATSTAT\\table_tls803.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls901_part01\\tls901_part01.csv TO D:\\PATSTAT\\table_tls901.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls902_part01\\tls902_part01.csv TO D:\\PATSTAT\\table_tls902.csv\n",
|
|
"MOVE: D:\\PATSTAT\\data_PATSTAT_Global_2022_Autumn_10\\tls904_part01\\tls904_part01.csv TO D:\\PATSTAT\\table_tls904.csv\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"file_path_dict=dict()\n",
|
|
"\n",
|
|
"# iterate over files in\n",
|
|
"# that directory\n",
|
|
"for root, dirs, files in os.walk(workdir_path):\n",
|
|
" for filename in files:\n",
|
|
" if filename.endswith(\".csv\") and not filename.startswith(\"table\"):\n",
|
|
" path= os.path.join(root, filename)\n",
|
|
" file_path_dict[filename] = path\n",
|
|
"\n",
|
|
"complete_file_set = set()\n",
|
|
"for fname in file_path_dict.keys():\n",
|
|
" complete_file_set.add(fname.split(\"_\")[0])\n",
|
|
"complete_file_set = sorted(complete_file_set)\n",
|
|
"\n",
|
|
"for complete_file in complete_file_set:\n",
|
|
" file_list = [file for file in file_path_dict.keys() if complete_file in file]\n",
|
|
"\n",
|
|
" outfile_path = os.path.join(workdir_path,\"table_\"+complete_file+\".csv\")\n",
|
|
" # print(outfile_path,file_list)\n",
|
|
"\n",
|
|
" if len(file_list)==1:\n",
|
|
" file_path = file_path_dict.get(file_list[0])\n",
|
|
" print(\"MOVE:\",file_path ,\"TO\",outfile_path)\n",
|
|
" shutil.move(file_path, outfile_path)\n",
|
|
" else:\n",
|
|
" print(\"CONCAT:\",file_list ,\"TO\",outfile_path)\n",
|
|
"\n",
|
|
"\n",
|
|
" CHUNK_SIZE = 50000\n",
|
|
" with_header=True\n",
|
|
" first_one = True\n",
|
|
" for csv_file_name in file_list:\n",
|
|
" csv_file_path = file_path_dict.get(csv_file_name)\n",
|
|
" chunk_container = pd.read_csv(csv_file_path, chunksize=CHUNK_SIZE,low_memory=False)\n",
|
|
" for chunk in tqdm(chunk_container):\n",
|
|
" chunk.to_csv(outfile_path, mode=\"a\", index=False, header=with_header)\n",
|
|
" with_header=False\n",
|
|
" first_one = False\n",
|
|
"\n",
|
|
" print(\"deleting\")\n",
|
|
" for csv_file_name in file_list:\n",
|
|
" csv_file_path = file_path_dict.get(csv_file_name)\n",
|
|
" print(csv_file_path)\n",
|
|
" os.remove(csv_file_path)"
|
|
],
|
|
"metadata": {
|
|
"collapsed": false
|
|
}
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 2
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython2",
|
|
"version": "2.7.6"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 0
|
|
}
|