cannabis_results / dev /get_all_results.py
keeganskeate's picture
latest-2024-08-11 (#6)
d1ae506 verified
"""
Get All Results | Cannabis Results
Copyright (c) 2024 Cannlytics
Authors:
Keegan Skeate <https://github.com/keeganskeate>
Created: 7/10/2024
Updated: 8/14/2024
License: <https://github.com/cannlytics/cannlytics/blob/main/LICENSE>
Description:
Aggregate cannabis test data from states with permitted cannabis use:
- Alaska
βœ“ California
- Colorado
βœ“ Connecticut
βœ“ Florida
βœ“ Hawaii
βœ“ Maryland
βœ“ Massachusetts
βœ“ Michigan
βœ“ Nevada
βœ“ New York
βœ“ Oregon
βœ“ Rhode Island
βœ“ Utah
βœ“ Washington
"""
# Standard imports:
import json
import os
# External imports:
from cannlytics.data.coas import standardize_results
from cannlytics.data.coas.parsing import find_unique_analytes
import pandas as pd
def get_all_results(
data_dir,
subsets=None,
ext='csv',
version='latest',
verbose=True,
):
"""Save all of the lab results to a single CSV file."""
# Read all of the latest datafiles.
all_data = []
for root, _, files in os.walk(data_dir):
for file in files:
if 'all' in file:
continue
if file.endswith(f'{version}.{ext}'):
file_path = os.path.join(root, file)
if verbose:
print(f"Reading {file_path}")
if ext == 'jsonl':
subset_data = pd.read_json(file_path, lines=True)
elif ext == 'csv':
try:
subset_data = pd.read_csv(file_path, low_memory=False)
except:
subset_data = pd.read_excel(file_path.replace('.csv', '.xlsx'))
else:
subset_data = pd.read_excel(file_path)
all_data.append(subset_data)
if verbose:
print(f"Read {len(subset_data):,} results from {file_path}")
# Merge historic MI results.
# TODO: Separate this logic into `get_results_mi.py`
# import ast
# from cannlytics.data.coas import CoADoc
# from cannlytics.utils import snake_case, convert_to_numeric
# datafile = r"D:\data\cannabis_results\data\mi\mi-results-psi-labs-2022-07-12.xlsx"
# psi_results = pd.read_excel(datafile)
# psi_results['lab_state'] = 'MI'
# psi_results['producer_state'] = 'MI'
# psi_results['date'] = pd.to_datetime(psi_results['date_tested'], format='mixed', errors='coerce')
# psi_results['week'] = psi_results['date'].dt.to_period('W').astype(str)
# psi_results['month'] = psi_results['date'].dt.to_period('M').astype(str)
# psi_results = psi_results.sort_values('date')
# Add a `key` to the PSI Labs results.
# parser = CoADoc()
# fixed_results = []
# for i, row in psi_results.iterrows():
# fixed_result = []
# sample_results = ast.literal_eval(row['results'])
# for result in sample_results:
# name = result['name']
# key = parser.analytes.get(snake_case(name), snake_case(name))
# value = convert_to_numeric(result['value'].replace('%', '').strip())
# fixed_result.append({
# 'key': key,
# 'name': name,
# 'value': value,
# 'margin_of_error': result.get('margin_of_error'),
# 'limit': result.get('limit'),
# 'status': result.get('status'),
# })
# fixed_results.append(json.dumps(fixed_result))
# psi_results['results'] = fixed_results
# analytes = find_unique_analytes(psi_results)
# nuisance_analytes = [
# '',
# '1_2_dimethoxy_ethane',
# '2_2_dimethyl_butane',
# '2_3_dimethyl_butane',
# '3_methyl_pentane',
# 'metrc_mi_cannabinoids_status',
# 'metrc_mi_cannabinoids_value',
# 'n_n_dimethylformamide',
# 'percent_active_cbd',
# # TODO: Standardize these analytes.
# # 'xylene'
# # 'hexane',
# # 'hexanes',
# # 'transnerolidol_1',
# # 'transnerolidol_2',
# ]
# analytes = set(analytes) - set(nuisance_analytes)
# analytes = sorted(list(analytes))
# psi_results = standardize_results(psi_results, analytes)
# outfile_csv = r'D:\data\cannabis_results\data\mi\mi-results-psi-labs-2022-07-12.csv'
# psi_results.to_csv(outfile_csv, index=False)
# print('Saved CSV:', outfile_csv)
# DEV: Merge historic PSI Labs results.
datafile = r"D:\data\cannabis_results\data\mi\mi-results-psi-labs-2022-07-12.xlsx"
psi_results = pd.read_excel(datafile)
all_data.append(psi_results)
if verbose:
print('Added %i historic lab results for PSI Labs.' % len(psi_results))
# Save all of the licenses.
aggregate = pd.concat(all_data)
outfile = os.path.join(data_dir, 'all/all-results-latest.csv')
aggregate.to_csv(outfile, index=False)
if verbose:
print('Aggregated %i results.' % len(aggregate))
print('Saved CSV:', outfile)
features = {x: 'string' for x in aggregate.columns}
print('Number of features:', len(features))
print('Features:', features)
return aggregate
# === Test ===
# [βœ“] Tested: 2024-08-14 by Keegan Skeate <keegan@cannlytics>
if __name__ == '__main__':
# Get all of the results.
get_all_results(data_dir='D://data/cannabis_results/data')