# Git + Obsidian

### Committing to Git Advice:
- **Commit Frequently**: It's a good practice to commit your changes often, even if they are small. This helps track the history of changes and makes it easier to roll back if needed.
- **Write Descriptive Commit Messages**: Ensure your commit messages clearly describe the changes made. This is especially important for tracking your progress or when collaborating with others. Use messages like:
  - `"Updated README with Obsidian setup instructions"`
  - `"Added meeting notes for 10/15/2024"`
  - `"Initial commit of thesis project folder structure"`
- **Use Git Push Regularly**: After committing, don't forget to push your changes to GitHub. This ensures your work is backed up remotely.

### Steps to Set Up GitHub for Obsidian Vault:

1. **Create a GitHub Repository:**
   - Go to GitHub, click on the **+** in the top right, and choose **New Repository**.
   - Name your repository `Thesis-Project-Vault`.
   - Add an optional description.
   - Choose **Private**.
   - **Do not initialize with a README** if you already have one in your Obsidian vault.
   - Click **Create repository**.

2. **Initialize Git in Your Local Vault:**
   - Open the terminal in the folder where your Obsidian vault is stored.
   - Initialize a Git repository:
     ```bash
     git init
     ```
   - Add your GitHub remote:
     - For HTTPS:
       ```bash
       git remote add origin https://github.com/username/Thesis-Project-Vault.git
       ```
     - For SSH:
       ```bash
       git remote add origin git@github.com:username/Thesis-Project-Vault.git
       ```

3. **Stage and Commit Your Files:**
   - Add and commit your files:
     ```bash
     git add .
     git commit -m "Initial commit"
     ```

4. **Push to GitHub:**
   - Push your commit to the GitHub repository:
     ```bash
     git push -u origin master
     ```

### Using a Personal Access Token (PAT) for Authentication:

1. **Generate a Token:**
   - Go to GitHub and navigate to **Settings** > **Developer Settings** > **Personal Access Tokens**.
   - Click **Generate new token**, and give it appropriate scopes (`repo` for repository access).
   - Copy the generated token.

2. **Set Git to Use HTTPS and Store Credentials:**
   - If using HTTPS, set the remote URL:
     ```bash
     git remote set-url origin https://github.com/username/Thesis-Project-Vault.git
     ```
   - When prompted for a username and password, use your **GitHub username** and the **Personal Access Token** as the password.

3. **Store Credentials Locally**:
   - To store the token so Git doesn’t ask every time:
     ```bash
     git config --global credential.helper store
     ```

### Installing and Configuring the Obsidian Git Plugin:

1. **Install the Plugin:**
   - Open **Obsidian**.
   - Go to **Settings** > **Community Plugins** > **Browse** and search for **Obsidian Git**.
   - Install and enable the plugin.

2. **Configure the Plugin:**
   - In **Settings**, scroll down to **Obsidian Git**.
   - Set the **Auto pull** and **Auto push** intervals if you want automatic syncing.
   - Test the configuration by using the command palette (`Ctrl + P` or `Cmd + P`) and searching for:
     - **Git: Commit all changes**
     - **Git: Push**
     - **Git: Pull**

### Testing and Verifying Setup:

1. **Verify on GitHub**:
   - After performing a push, check your GitHub repository to ensure the files have been successfully uploaded.

2. **Sync with Git in Obsidian**:
   - Use the **Command Palette** in Obsidian to manually commit, push, or pull changes as needed.

3. **Optional: Developer Console for Logs**:
   - Open **Developer Tools** (`Ctrl + Shift + I`) in Obsidian to check for any errors or logs related to Git operations.

### Conclusion:
By following these steps, you have successfully configured your Obsidian vault with Git and GitHub for version control. Regular commits and pushes will ensure that your work is backed up and easy to track over time.

# Example of linking specific paper notes as text for brainstorming. 
### **Finding Satellite-Halo Mass Relations**

The relationship between ultra-diffuse galaxy (UDG) numbers and halo mass appears to hold even for fainter galaxies than previously thought. [[zaritskyLIGHTSSurveyOverview2024#^annotation-16|Recent work found only 0.11 ± 0.11 UDG satellites per host]] in galaxies roughly 2.5 times fainter than Milky Way analogues, which is consistent with theoretical predictions of 0.07 ± 0.01 UDGs expected from halo mass scaling relations.

### **Observational Depth Advantages**

What sets the zaritskyLIGHTSSurveyOverview2024 apart is its observational depth. [[zaritskyLIGHTSSurveyOverview2024#^annotation-18|The survey achieves a 3σ surface brightness limit of ~30.5 mag arcsec⁻² in r-band, reaching sources approximately 2 magnitudes fainter than Legacy Survey images]]. This deeper sensitivity has revealed [[zaritskyLIGHTSSurveyOverview2024#^annotation-7|improved astrometric solutions and photometric calibration through multi-step refinement processes]]. [[zaritskyLIGHTSSurveyOverview2024]] 

### **Methodological Considerations**

The flat-field correction methodology is crucial for deep imaging. [[zaritskyLIGHTSSurveyOverview2024#^annotation-6|Rather than using traditional dome flats, the team developed a nightly master flat creation process using science images, with strict rejection criteria for contaminating sources]]. This approach reduces systematic errors that plague shallower surveys.

### **Comparative Results**

Interestingly, [[zaritskyLIGHTSSurveyOverview2024#^annotation-17|nuclear star cluster occurrence rates in LSB satellites match those found in MATLAS dwarfs (22-23%)]], suggesting that host galaxy morphology may not significantly influence NSC formation in satellite systems.

# Basic markdown formatting 

`# Heading 1
`## Heading 2
`### Heading 3

**bold text**
*italic text*
- bullet point
  - nested bullet

[[Link to Another Note]]
[[Another Note#Specific Section]]
[[Note#^annotation-5|custom link text]]

#tag #research #astrophysics
```

**Creating and Linking Notes**
- Type `[[` and start typing a note name
- Obsidian auto-suggests existing notes
- If it doesn't exist, hitting Enter creates it
- This is how you build your knowledge graph

**Daily Notes (Optional but useful)**
- Press Cmd+P, type "Daily note", hit Enter
- Creates a note dated today
- Great for research journals
- Can be templated

**Opening Multiple Panes**
- Click and drag tabs to split screen
- Reference one note while writing another
- Super useful when synthesizing papers

---

## **Connecting Zotero to Obsidian**
**Title:** Connect Zotero & Obsidian

**Text Instructions:**

1. **Tell Obsidian Where Your Zotero Library Is**
   - Open Settings > Citations (left sidebar)
   - Find: "Citation database format" - select "JSON"
   - Find: "Citation export path" - click the folder icon
   - Navigate to your Zotero data folder (usually in Documents/Zotero)
   - Select the folder and confirm

2. **Set Up Your Literature Note Template**
   - In Settings > Citations, find "Literature note content template"
   - Paste the hybrid template we created (provided separately)
   - This controls how your papers appear when imported

3. **Create a Hotkey for Quick Import**
   - In Settings, search for "hotkeys"
   - Search for "Citations: Create literature note"
   - Click and press a shortcut (e.g., Cmd+Shift+Z or Ctrl+Shift+Z)
   - Now you can import papers with one keystroke!

4. **Test It Out**
   - Open any note in Obsidian
   - Press your hotkey
   - A popup appears showing your Zotero library
   - Search for a paper you annotated
   - Click it → Literature note automatically imports with all your highlights

**What You'll See:**
- All your Zotero highlights in colored callouts
- Block identifiers (^annotation-1, ^annotation-2, etc.)
- Each annotation links back to the exact PDF page
- Tags and comments from Zotero also import

---

## **Syncing with Git**
**Title:** Keep Everything in Sync

**Text Instructions:**

1. **Configure Obsidian Git**
   - Open Settings > Obsidian Git
   - Leave most settings as default
   - Find "Commit message on interval" and enable it
   - This auto-commits every hour (or set custom interval)

2. **Manual Sync (Do This Every Session)**
   - Press Cmd+P (or Ctrl+P)
   - Type "Obsidian Git: Pull"
   - This fetches latest notes from GitHub
   - Work on your notes locally
   - Type "Obsidian Git: Commit and Push"
   - This uploads all changes to GitHub

3. **Quick Git Commands (If You Prefer Terminal)**

   cd path/to/your/vault
   git pull                # Get latest version
   git add .               # Stage all changes
   git commit -m "Added notes on dark matter models"
   git push                # Upload to GitHub
```
# Local AI in Obsidian — Setup Guide

> **What this gives you:** A free, private AI assistant living inside Obsidian. No subscriptions, no data sent to the cloud. Everything runs on your own machine.

---

## What You Need to Install

1. **Ollama** — runs the AI model locally on your computer
2. **A model** — the actual AI brain (we'll download one)
3. **Smart Composer** — the Obsidian plugin that connects everything

---

## Part 1: Install Ollama

### Mac

```bash
brew install ollama
```

> Don't have Homebrew? Install it first from [brew.sh](https://brew.sh/), or download Ollama directly from [ollama.com/download](https://ollama.com/download).

### Windows / Linux

Download the installer directly from [ollama.com/download](https://ollama.com/download) and follow the on-screen steps.

---

## Part 2: Start Ollama & Download a Model

Open your Terminal (Mac/Linux) or Command Prompt (Windows) and run:

```bash
ollama serve
```

Leave that window open. Then open a **second terminal window** and run:

```bash
ollama pull qwen3:4b
```

This downloads the AI model (~2.5 GB). 

> **Want a smaller/faster model for testing?** `ollama pull qwen3:0.6b` is only ~400 MB and downloads in under a minute.

> **Have 16 GB RAM and want something smarter?** `ollama pull qwen3.5:9b` (~5 GB) is noticeably more capable for coding and analysis.

---

## Part 3: Install Smart Composer in Obsidian

1. Open Obsidian
2. Go to **Settings → Community Plugins**
3. Click **Browse** and search for **Smart Composer**
4. Click **Install**, then **Enable**

---

## Part 4: Connect Smart Composer to Ollama

1. Go to **Settings → Smart Composer**
2. Scroll down to **Providers → Ollama** and click the ⚙️ gear icon
3. Confirm the Base URL is: `http://localhost:11434`

**Register your model:** 4. Scroll to **Chat Models → Add Custom Model** - ID: `qwen3:4b` - Provider: `ollama` - Model Name: `qwen3:4b` 5. Click **Add** and make sure the toggle is **ON**

**Activate it:** 6. Scroll up to **General** - Chat Model → `qwen3:4b` - Apply Model → `qwen3:4b`

---

## Part 5: Test It

Open any note in Obsidian and press `Cmd+L` (Mac) or `Ctrl+L` (Windows/Linux) to open the AI chat sidebar. Type something and hit Enter. You should get a response within a few seconds.

---

## Tips & Common Questions

**Do I need to keep Ollama running all day?** Yes, but it's lightweight when idle (~50–100 MB RAM). The model itself only loads into memory when you're actively chatting, then unloads automatically after 5 minutes of inactivity.

**Can I turn it off when not using it?** Yes — just close the terminal window running `ollama serve`. Reopen it when you need AI again.

**The model is slow on my machine — what should I do?** Switch to a smaller model like `qwen3:0.6b` or `qwen3:1.7b`. Smaller = faster, just a bit less capable.

**I get a "connection refused" error:** Ollama isn't running. Open a terminal and run `ollama serve`.

**How much RAM do I need?**

|RAM|Recommended Model|
|---|---|
|8 GB|`qwen3:4b`|
|16 GB|`qwen3:4b` or `qwen3.5:9b`|
|32 GB+|`qwen3:14b` or larger|

---

## Uninstalling (If Needed)

```bash
# Remove downloaded models
rm -rf ~/.ollama

# Uninstall Ollama (Mac with Homebrew)
brew uninstall ollama

# Uninstall Ollama (Mac direct install)
sudo rm /usr/local/bin/ollama
```

To remove Smart Composer: **Settings → Community Plugins → Smart Composer → Uninstall**

Your Obsidian vault and notes are never touched by any of this.

---

_Guide based on setup tested April 2026 on macOS with Ollama 0.21+_