Rev Author Line No. Line
4928 kaklik 1 #!/usr/bin/python
2  
4929 kaklik 3 import pandas as pd
4 import os
5 import time
6 import datetime
4928 kaklik 7  
4929 kaklik 8 from pymlab import config
9 from mlabutils import ejson
4928 kaklik 10  
4929 kaklik 11 parser = ejson.Parser()
4928 kaklik 12  
4929 kaklik 13 #### Script Arguments ###############################################
4928 kaklik 14  
4929 kaklik 15 if len(sys.argv) != 2:
16 sys.stderr.write("Invalid number of arguments.\n")
17 sys.stderr.write("Usage: %s CONFIG_FILE\n" % (sys.argv[0], ))
18 sys.exit(1)
19  
20 value = parser.parse_file(sys.argv[1])
21 dataSource = value['data_path']
22 dataArchive = value['data_archive']
4930 kaklik 23 dataUpload = value['data_upload']
24 stationName = value['origin']
25  
4929 kaklik 26 loop = 1
27  
28  
4928 kaklik 29 while True:
4929 kaklik 30 try:
4928 kaklik 31 print("Start")
32 ## Create sorted list of csv files
33 listOfDataFiles = list() #empty list
34 listOfSpecDataFiles = list() #empty list
35 files = list() #empty list
36 falg = False # is computation needed
4929 kaklik 37  
38 files = sorted(os.listdir(dataSource)) # list of all files and folders in directory
39 for idx, val in enumerate(files): #goes through files
40 if val.endswith("data.csv"): # in case of *data.csv
41 listOfDataFiles.append(val) #add file to listOfFiles
42  
43 ## Find the newest and oldest and compare them. If they are from different day, compute the average of all measurement from oldest day
44 if len(listOfDataFiles)>=2: # if there are more than 2 data files
45 first = listOfDataFiles[0] # get first of them
46 last = listOfDataFiles[-1] # get last of them
47  
48 if time.mktime(datetime.datetime.strptime(last[:8], "%Y%m%d").timetuple()) > time.mktime(datetime.datetime.strptime(first[:8], "%Y%m%d").timetuple()): # if the last is older than first
4928 kaklik 49 flag = True # computation needed
50 print("Computing...")
51 print(loop)
4929 kaklik 52 loop +=1
53 listOfSpecDataFiles = list() # empty list
54  
55 for file in listOfDataFiles: # go through data files and create lis of data files measured on same day
56 # if the day is same like the first one
57 if time.mktime(datetime.datetime.strptime(first[:8], "%Y%m%d").timetuple()) == time.mktime(datetime.datetime.strptime(file[:8], "%Y%m%d").timetuple()):
58 listOfSpecDataFiles.append(file)
59  
60 for file in listOfSpecDataFiles:
61 df=pd.read_csv(file, sep=';', header=None) # read current csv
62 dim=df.shape # gets data file dimensions
63 rowsInd=dim[0] # maximal index of rows
64 columnsInd=dim[1] # maximal index of columns
65 values=pd.DataFrame() # empty DataFrame
66  
67 for x in range(0,columnsInd): # for each column
68 values = values.set_value(0,x,round(df[x].mean(),3),0) #calculates mean value for all cloumns and round it by 3
4928 kaklik 69  
4930 kaklik 70 filename = dataUpload + first[:8]+'000000_' + stationName + '_data_mean.csv'
4928 kaklik 71 outfile = open(filename, 'a')
4929 kaklik 72 values.to_csv(filename, sep=';', header=None, index=False, mode='a') # save (add) DataFrame to csv
4928 kaklik 73 outfile.close()
74  
75 # move files to archive structure
76 for file in listOfSpecDataFiles:
77 year = file[:4]
78 month = file[4:6]
79 day = file[6:8]
80 directory = dataArchive + year + "/" + month + "/" + day + "/"
81 if not os.path.exists(directory):
82 os.makedirs(directory)
83 os.rename(dataSource + file, dataArchive + year + "/" + month + "/" + day + "/" + file) # move file
84  
85 else:
86 flag = False # computation is not needed
87 else:
88 flag = False # computation is not needed
89  
90 if flag == False:
91 time.sleep(10) #long sleep, because is nothing to process
92  
93 except ValueError:
94 print ValueError