playground
Learn PyCharm by clicking through it
A simulated PyCharm 2026.2 window, set up the way a data engineer works: a uv project, a Polars pipeline, a notebook and a Postgres warehouse. Follow a guided tour and watch each step happen, or click any part of the window to find out what it is for. Then drill the shortcuts with your real keyboard.
Step 1 of 8
Open a repo, give it a uv interpreter, stop it indexing your data. Click any part of the window to ask what it is. ← → step through.
A simulation matching PyCharm 2026.2 on macOS with the default keymap. Output and timings are illustrative. Nothing leaves your browser.
first_hour
Your first hour on a data repo
In the order that saves the most pain. None of it takes more than two minutes.
- 1
Give the project its own interpreter
A uv environment in .venv, created from uv.lock. Red imports on code that runs fine in your terminal almost always mean the wrong interpreter.
Status bar → interpreter widget → Add New Interpreter
- 2
Exclude every data folder
CSV, Parquet, dbt target/ and notebook checkpoints. Indexing them is why a data repo feels slow, and why searches return rows of data instead of code.
Right-click folder → Mark Directory as → Excluded
- 3
Switch on Ruff, and one type checker
Native since 2025.3, with no plugin needed. Pick one type checker (ty, Pyright or Pyrefly) to avoid duplicate warnings, and commit .idea/pyLspTools.xml so the team matches.
Settings → Python → Tools → External Tools
- 4
Put credentials in .env files, not your shell
Each run configuration can load one or more .env files. The pipeline then runs the same way from the IDE, the terminal and CI.
Run → Edit Configurations… → Paths to .env files
- 5
Connect the warehouse read-only
Add the data source with a read-only user and tick Read-only as well. That's two locks on the one door you can't afford to leave open.
Database → + → Data Source → Options → Read-only
- 6
Give it more memory
The default heap is sized for small projects. On a monorepo, 4096 MB stops the pauses while it indexes.
Help → Change Memory Settings
- 7
Install Key Promoter X
Every time you click something that has a shortcut, it tells you the shortcut. It's the fastest way off the mouse.
Settings → Plugins → Marketplace
- 8
Back up your setup
Keymap, plugins and settings follow you to the next laptop or a reinstall.
File → Manage IDE Settings → Backup and Sync
plugins
Plugins worth knowing, and three to avoid
Everything lives in Settings → Plugins. Bundled ones are already installed and can be switched off. Marketplace ones are one click to add. Fewer plugins means a faster IDE, so add them when you need them, not all at once.
| Plugin | Status | Why a data engineer cares |
|---|---|---|
| Jupyter | Bundled | Native .ipynb editing, interactive DataFrame tables, the variables window. Free tier. |
| Database Tools and SQLPro | Bundled | The DataGrip engine: data sources, schema-aware SQL completion, an editable results grid, export. Postgres, Snowflake, BigQuery, Redshift, ClickHouse and more. |
| DockerPro | Bundled | Containers in the Services window, and a container as the project interpreter, so the IDE runs exactly what production runs. |
| Remote development (SSH)Pro | Bundled | An SSH or remote interpreter for code that has to run next to the data, on a box with the memory for it. |
| AI Assistant, Junie and ACP agents | Bundled | Chat and agents in the IDE, including Claude Code or Codex over the Agent Client Protocol. Needs a JetBrains AI plan or your own agent's subscription. |
| Big Data Tools | Marketplace | Spark job submission and monitoring, Kafka topics and schema registry, and S3, GCS and Azure storage browsers. If it won't load on 2026.2, update the plugin; that was a known bug (IJPL-249064). |
| AWS Toolkit | Marketplace | S3 browsing, Lambda, CloudWatch logs and credentials profiles, without switching to the console. |
| Rainbow CSV | Marketplace | Colours each CSV column so you can read a 40-column vendor extract without a spreadsheet. |
| .env files | Marketplace | Syntax, completion and go-to for environment variables across .env files and your code. |
| Key Promoter X | Marketplace | Shows the shortcut for whatever you just clicked. Uninstall it once it stops popping up. |
| Data Wrangler, Hugging Face, Google Colab | Retiring | Unbundled in 2026.2 as low-usage. JetBrains stops publishing compatible versions from 2026.3, so don't build a workflow on them. |
free_or_pro
What needs a subscription
Since 2025.1 there is one PyCharm, with Community and Professional merged into it. Every new install starts a 30-day Pro trial, so you'll see Pro features first and lose them later if you don't subscribe.
Free
Editor, refactoring, debugger, Git, Terminal, Python Console, Jupyter notebooks, uv/Poetry/Conda interpreters, Ruff and type checkers, and JavaScript/TypeScript since 2026.1.
Pro
Database tools and SQL, Docker and SSH interpreters, remote development, the profiler, the HTTP Client, and Django/Flask/FastAPI support. AI features need a separate JetBrains AI plan.
faq
Questions & answers
- Is PyCharm free?
- Since 2025.1 there is one PyCharm with a free core and a Pro subscription on top. The editor, debugger, Git, terminal and Jupyter notebooks are free. Database tools, Docker and SSH interpreters, the profiler and web framework support need Pro, and every new install starts with a 30-day Pro trial.
- What are the most important PyCharm shortcuts on a Mac?
- Search Everywhere (tap Shift twice), Find Action (⇧⌘A), Recent Files (⌘E), Go to Definition (⌘B), Show Context Actions (⌥↩), Rename (⇧F6), Run (⌃R) and Debug (⌃D). If you only learn one, learn Find Action: it finds every other command and shows its shortcut.
- How do I use uv with PyCharm?
- Click the interpreter widget in the status bar, choose Add New Interpreter → Add Local Interpreter, and pick uv as the type. It has been the default since 2025.3. PyCharm creates .venv next to pyproject.toml and installs from uv.lock. Add dependencies with uv add in the Terminal so the lock file stays the source of truth.
- Why does PyCharm say my imports are unresolved when the code runs fine?
- The project is pointed at a different interpreter than the one your terminal uses. Check the interpreter widget in the status bar. If it isn't the project's .venv, switch it there. If it is, File → Invalidate Caches and a restart clears a stale index.
- Can I use Jupyter notebooks in PyCharm?
- Yes, and it is part of the free tier. .ipynb files open natively, ⇧↩ runs a cell, DataFrames render as sortable tables, and the Jupyter Variables window shows what the kernel is holding. Since 2026.2.1, AI agents can also edit and run cells against the live kernel.
- Can PyCharm connect to Postgres, Snowflake or BigQuery?
- With Pro, yes. The Database tool window is the DataGrip engine, with data sources for Postgres, MySQL, Snowflake, BigQuery, Redshift, ClickHouse and many more, SQL completion against the real schema, and an editable results grid. Tick Read-only on production data sources.
- Why is PyCharm slow on my data repository?
- Usually it is indexing your data. Right-click folders full of CSV, Parquet or dbt build output and choose Mark Directory as → Excluded. On a large monorepo, also raise the heap under Help → Change Memory Settings.
- Which plugins should a data engineer install?
- Start with Key Promoter X to learn shortcuts, a CSV plugin such as Rainbow CSV, and the .env files plugin. Add Big Data Tools (Spark, Kafka, cloud storage) and the AWS Toolkit if you use them. Ruff and ty are supported natively since 2025.3, so they no longer need a plugin.
- Is this the real PyCharm?
- No. It's a simulation that runs in your browser, built to match PyCharm 2026.2 on macOS with the default keymap. Layouts vary a little between versions and with your settings, and output and timings are illustrative. Nothing you type or click leaves the page.