Python Basics for Beginners: Install Python 3.14.7 to CSV Summary

As of September 30, 2026, install Python by downloading the latest stable release, 3.14.7 (released August 5), from python.org. On Windows 10 or later, one line with the Python Install Manager finishes the job: py install 3.14 (per the official Windows docs). After installing, check the version with py --version. Then summarize an ad report CSV with a 15-line script, and you will use five core Python syntax features in one go. This guide follows that exact order.

The short version in three lines

• Version to download: 3.14.7. Support for 3.10 ends in October 2026.
• Windows install: starting with 3.14 (released October 7, 2025), the official tool is the Python Install Manager. The old "full installer" and "Python launcher" are marked as deprecated.
• First run: create a virtual environment (.venv) for each project folder, then calculate CPA per campaign using only the standard library csv module.

The sources I checked are the python.org download page and the official Python docs for Windows usage, the virtual environment tutorial, and the csv module. I ran the example script on a sample CSV to confirm the output format. This guide covers Windows 10 or later only. macOS and Linux are not covered. I could not reproduce the Windows installer screens myself, so the commands and troubleshooting items follow the official docs.

The Python to download now is 3.14.7

The python.org download page lists 3.14.7 as the latest stable release. It came out on August 5, 2026. On the same page, 3.14 and 3.13 are in the bugfix stage, and 3.12 and earlier are in the security-fix stage. Python 3.10 is listed as reaching end of support in October 2026, and the page gives only the month, not an exact date. If your PC runs 3.10, MCPY thinks you should upgrade this month (MCPY recommendation).

If you are just starting out, download 3.14 (MCPY recommendation). If your company tools or libraries state that they support only a specific version, use that version first. MCPY left Anaconda and the free-threaded build (3.14t) off the first-install list. You do not need them to learn the basics, and more choices at the start means more places to get stuck.

On Windows, three commands finish the install

The Python Install Manager is a tool that downloads and manages the Python runtime (the program that actually executes your code). Per the official docs, Python 3.14 is supported on Windows 10 or later. Get the install manager from the Microsoft Store or python.org, then type the following in a terminal (Command Prompt or PowerShell), in order.

1. py install 3.14 installs the Python 3.14 runtime.
2. py --version checks the installed version.
3. python hello.py runs a hello.py file you created in Notepad.

If you use winget, you can also get the install manager with winget install 9NQ7512CXL7T -e (this is the package ID given in the official docs). py list shows the runtimes currently installed.

Only one line is needed inside hello.py.

print("안녕, 파이썬")

When the python command fails, check three things

Here is the official troubleshooting guidance, condensed.

• python command not found: confirm the install manager is installed. Then open "Manage app execution aliases" in Windows Settings and check that the Python-related entries are turned on.
• The wrong version runs: check the value of the PYTHON_MANAGER_DEFAULT environment variable, and check whether an older Python install is still present.
• Conflict with the old py.exe: find "Python launcher" in the installed apps list and uninstall it.

If the installer asks about PATH (the list that decides which folders' programs you can call directly by name), MCPY recommends saying yes to adding it. The official docs say this option is offered at the first runtime install. Per the official docs, the default location is %LocalAppData%\Python\bin.

Create one virtual environment per project

A virtual environment is a folder that holds each project's libraries separately. One project can pin requests 2.6.0 while another uses the latest version, and they never collide. The steps from the official tutorial are below.

1. python -m venv .venv creates .venv in the current folder.
2. .venv\Scripts\activate activates it. If (.venv) appears at the start of your prompt, it worked. In PowerShell, use .venv\Scripts\Activate.ps1.
3. python -m pip install requests installs the library you need.

To rebuild the same environment on another PC later, save the list with python -m pip freeze > requirements.txt. Then run python -m pip install -r requirements.txt on the new PC. When you are done, turn the environment off with deactivate. If PowerShell says script execution is blocked, open Command Prompt and try again. This tip is not in the official docs. It is a common case, and I could not verify it myself.

A first script that summarizes an ad report CSV

Reading one useful script is faster than memorizing basic syntax on its own. Say the ad_report.csv you downloaded from an ad manager looks like the sample below. The top line holds the column names. (In this file, 캠페인 means campaign, 비용 means cost, 클릭 means clicks, and 전환 means conversions.)

캠페인,비용,클릭,전환
브랜드검색,120000,340,12
쇼핑,250000,610,20
브랜드검색,80000,210,5
리타겟팅,90000,150,0

Create report.py in the same folder and paste in the following. If a campaign appears on several rows, it adds them up and calculates CPA (cost per acquisition, the cost of getting one conversion).

import csv
from collections import defaultdict

cost = defaultdict(int)
conv = defaultdict(int)

with open("ad_report.csv", encoding="utf-8-sig", newline="") as f:
    for row in csv.DictReader(f):
        name = row["캠페인"]
        cost[name] += int(row["비용"])
        conv[name] += int(row["전환"])

for name in cost:
    cpa = cost[name] / conv[name] if conv[name] else 0
    print(f"{name}: 비용 {cost[name]:,}원, 전환 {conv[name]}건, CPA {cpa:,.0f}원")

When you run python report.py, the sample produces this output.

브랜드검색: 비용 200,000원, 전환 17건, CPA 11,765원
쇼핑: 비용 250,000원, 전환 20건, CPA 12,500원
리타겟팅: 비용 90,000원, 전환 0건, CPA 0원

The basic syntax used in the code

• Variables: names attached to values, like name and cpa. The = sign means "store", not "equals".
• Dictionaries: boxes where you look up a value by a label (a key), like cost[name]. defaultdict(int) makes a key you have not seen before start at 0.
• Loops: for row in ... takes each CSV row one at a time and runs the indented lines below it. Indentation is part of the syntax. If the number of spaces is off, you get an IndentationError.
• Conditional expressions: A if condition else B avoids the divide-by-zero error (ZeroDivisionError) when there are 0 conversions.
• f-strings: the f before the string and the curly braces {} insert values. :, adds thousands separators.

The csv module docs say it is always safe to specify newline='' when opening the file. The code above does that. You must change the column names 캠페인, 비용, and 전환 to match the first line of your own file. If they differ, you get a KeyError.

What to check if you work in Korea

• Download sites: searching for "파이썬 다운로드" (Python download) also returns sites that offer the installer on their own. Check first that the address bar says python.org.
• CSV encoding: it is commonly reported that Excel on Korean Windows saves "CSV (comma delimited)" as cp949, a legacy Korean encoding. (This paragraph is general knowledge, not from the official docs, and MCPY could not verify it directly.) If the code above shows garbled characters or a UnicodeDecodeError, change it to encoding="cp949". A file saved as "CSV UTF-8" opens with utf-8-sig. This varies by file, and MCPY could not check every case.
• Column names: the CSV files from the Naver (Korea's largest search engine), Kakao, and Google ad managers use different column names and orders. For each platform, edit row["..."] in the script to fit that file.

Common mistakes

• Running pip install without activating the virtual environment. The library lands in the global install, and versions get mixed up later. First check that (.venv) shows in your prompt.
• Using pip instead of python -m pip, which can install into a different Python. The official tutorial also uses the python -m pip form.
• Naming your file csv.py. It clashes with the standard module, and import csv loads your own file. Name it something like report.py.
• Only reading the example with your eyes (MCPY recommendation: type it yourself). Deliberately break one column name and see the KeyError yourself. That is how you learn to read error messages.

What to do today

This week, run py install 3.14. Then download last week's report from your own ad account as a CSV, change only the column names in report.py, and run it. Once you get your first result, the next step is saving the output back to a CSV. Install commands and versions change often, so check the latest version on the python.org download page once more before you run anything.

Official sources referenced

• Python downloads (python.org)
• Using Python on Windows (official docs)
• Virtual environments and packages (official tutorial)
• csv module (official docs)