Mercurial > repos > jay > gaiac_precision

comparison gaiac_time_sync/gaiac_time_sync.py @ 0:287d6cc86582 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/jaidevjoshi83/gaiac.git commit c29a769ed165f313a6410925be24f776652a9663-dirty
author jay
date Thu, 15 May 2025 14:45:45 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:287d6cc86582
1 import pandas as pd
2 import os
3 import argparse
4
5 def align_sensor_data(file_list, date_time, sep=',', output_mode='multiple', output='aligned.tsv'):
6
7 file_list = file_list.split(',')
8 if len(file_list) < 2:
9 print("Please provide at least two files.")
10 return
11
12 # Read all files into a list of dataframes
13 dfs = [pd.read_csv(file, sep=sep, parse_dates=[date_time]) for file in file_list]
14
15 # Get common timestamps by successive inner merges
16 common_times = dfs[0][[date_time]]
17 for df in dfs[1:]:
18 common_times = common_times.merge(df[[date_time]], on=date_time, how='inner')
19
20 # Now filter each dataframe to contain only common timestamps
21 aligned_dfs = [
22 df[df[date_time].isin(common_times[date_time])].reset_index(drop=True)
23 for df in dfs
24 ]
25
26 # Output files
27 if output_mode == 'multiple':
28 for i, (file, df) in enumerate(zip(file_list, aligned_dfs)):
29 filename = os.path.splitext(os.path.basename(file))[0]
30 output_file = f"{output_prefix}_{filename}.csv"
31 df.to_csv(output_file, index=False, sep=sep)
32 print("Aligned files saved individually.")
33 elif output_mode == 'single':
34 # Merge all aligned dataframes on date_time
35 merged_df = aligned_dfs[0]
36 for df in aligned_dfs[1:]:
37 merged_df = merged_df.merge(df, on=date_time, how='inner')
38
39
40 merged_df.to_csv(output, index=False, sep=sep)
41 print("Single merged file saved.")
42 else:
43 print("Invalid output mode. Use 'multiple' or 'single'.")
44
45
46 def main():
47 parser = argparse.ArgumentParser(description="Align sensor data files on common timestamps.")
48
49 parser.add_argument(
50 '-f', '--files',
51 required=True,
52 help='List of input file paths (at least two)'
53 )
54
55 parser.add_argument(
56 '-s', '--sep',
57 default='\t',
58 help='Separator used in the input files (default: ,)'
59 )
60
61 parser.add_argument(
62 '-m', '--mode',
63 choices=['multiple', 'single'],
64 default='single',
65 help="Output mode: 'multiple' for individual files, 'single' for one merged file (default: multiple)"
66 )
67
68 parser.add_argument(
69 '-o', '--output',
70 default='aligned',
71 help="Output filename"
72 )
73
74 parser.add_argument(
75 '-t', '--date_time_column',
76 default='date_time',
77 help="Provide the name of the date and time column."
78 )
79
80 args = parser.parse_args()
81
82 align_sensor_data(
83 file_list=args.files,
84 date_time=args.date_time_column,
85 sep=args.sep,
86 output_mode=args.mode,
87 output=args.output
88 )
89
90 if __name__ == '__main__':
91 main()