프로젝트 종료일 예측

estimate.py

import requests
from requests.auth import HTTPBasicAuth
import pandas as pd
from datetime import datetime, timedelta
import numpy as np
import warnings
warnings.filterwarnings("ignore")

# Authentication for Jira API
USERNAME = ""  # Put your Jira username here.
API_TOKEN = ""  # Put your Jira API token here.
PROJECT = ""  # Put your project key here.
DOMAIN = ""  # Put your Jira domain here.
ISSUE_GROW_FACTOR = 1.2  # Adjust this factor as per your requirement.

# Jira API URL
url = f"https://{DOMAIN}.atlassian.net/rest/api/3/search"  # Replace '[your_domain]' with your actual domain.

# Headers
headers = {"Accept": "application/json"}

# Query
query = {
    "jql": f"project = '{PROJECT}' AND (resolution = Unresolved or resolved > startOfDay(-28d))",
    "maxResults": 1000,  # Maximum issues per request. Modify as per your needs.
    "fields": ["key", "resolutiondate"]  # Modify these fields as per your needs.
}

# Send a request to Jira API
response = requests.get(url, headers=headers, params=query, auth=HTTPBasicAuth(USERNAME, API_TOKEN))

# Convert the response to JSON
data = response.json()

# Create a DataFrame from the issues
issues = pd.json_normalize(data['issues'])

# We have 'fields.created', 'fields.updated' and 'fields.resolved' in our dataframe. 
# We will just consider 'fields.resolved' for your case. 
issues['fields.resolutiondate'] = pd.to_datetime(issues['fields.resolutiondate'])

# Count unresolved issues
remaining_tasks = issues['fields.resolutiondate'].isna().sum()
remaining_tasks = remaining_tasks * ISSUE_GROW_FACTOR

recent_data = issues.copy()

# Group by week and count the tasks completed
recent_data.loc[:, 'week'] = recent_data['fields.resolutiondate'].dt.to_period('W')
recent_tasks_completed = recent_data.groupby('week').size().values


# Monte Carlo simulation
num_simulations = 1000  # Modify this number as per your needs.
simulation_results = []

for _ in range(num_simulations):
    total_weeks = 0
    remaining_tasks_simulation = remaining_tasks
    while remaining_tasks_simulation > 0:
        tasks_per_week = np.random.choice(recent_tasks_completed)
        remaining_tasks_simulation -= tasks_per_week
        total_weeks += 1
    simulation_results.append(total_weeks)

# Convert simulation results to dates
simulation_end_dates = [datetime.now() + timedelta(weeks=w) for w in simulation_results]

# Convert to DataFrame
dates_df = pd.DataFrame(simulation_end_dates, columns=['Date'])

# Count the number of each date and sort by date
cumulative_counts = dates_df['Date'].dt.date.value_counts().sort_index().cumsum()

# Calculate the cumulative percentages
cumulative_counts = cumulative_counts / len(simulation_end_dates) * 100

# Print the cumulative counts
for date, percentage in cumulative_counts.iteritems():
    print(f'{date},{percentage}')

$ brew install youplot
$ python estimate.py | uplot bar -d, 

This will draw a cummulative histo-diagram of the project's estimated completion date.