merge_result

#! /usr/bin/env python

import pickle
import os
import csv
import argparse

def parse_args():
    # TODO add arg for globus refresh token
    description = 'usage '
    parser = argparse.ArgumentParser(description=description)
    parser.add_argument('-f', '--pkl-folder', help='path to the root of pkl', required=True)
    parser.add_argument('-o', '--output-path', help='/path/to/result.csv', default='./result.csv')
    args = parser.parse_args()
    return args


args = parse_args()

csv_path = args.output_path
pkl_path = os.path.join(args.pkl_folder, '')
columns = ['local_datetime', 'leaf_length', 'leaf_width', 'access_level', 'species',
           'site', 'citation_author', 'citation_year', 'citation_title', 'method']
access_level = 2
species = 'Sorghum bicolor'
citation_author = 'Zeyu Zhang'
citation_year = '2019'
citation_title = ''
method = ''
pkl_filename_list = os.listdir(pkl_path)
'''
merged_dict:
date::dict
    |- col, range::list
        |- leaf_length_sum
        |- leaf_width_sum
        |- leaf_count
'''
merged_dict = {}
for r_filename in pkl_filename_list:
    try:
        with open(os.path.join(pkl_path, r_filename), 'rb') as f:
            result = pickle.load(f)
    except:
        continue
    leaf_date = r_filename[:10]
    sensor_name = r_filename[-8:-4]
    if (leaf_date, sensor_name) not in merged_dict:
        merged_dict[(leaf_date, sensor_name)] = {}
    for leaf_width, leaf_length, col_range in \
            zip(result['leaf_width'], result['leaf_length'], result['col_range']):
        if leaf_width is None or leaf_length is None or col_range is None:
            continue
        if col_range not in merged_dict[(leaf_date, sensor_name)]:
            merged_dict[(leaf_date, sensor_name)][col_range] = [0, 0, 0]
        merged_dict[(leaf_date, sensor_name)][col_range][0] += leaf_length
        merged_dict[(leaf_date, sensor_name)][col_range][1] += leaf_width
        merged_dict[(leaf_date, sensor_name)][col_range][2] += 1

result_list = []
for date_sensor, date_sensor_dict in merged_dict.items():
    for col_range, col_range_data in merged_dict[date_sensor].items():
        row = [date_sensor[0],  # date
               date_sensor[1],  # sensor name
               col_range[0],       # col
               col_range[1],
               '{:.3f}'.format(col_range_data[0]/col_range_data[2]),  # average leaf_length
               '{:.3f}'.format(col_range_data[1]/col_range_data[2])]  # average leaf width
        result_list.append(row)

with open(csv_path, 'w') as csvfile:
    csv_writer = csv.writer(csvfile)
    csv_writer.writerows(result_list)