update

.gitignore

@@ -1,2 +1,5 @@
 /.venv/
-.venv
+.venv
+/.vscode/
+.vscode
+.obsidian

.vscode/launch.json

@@ -1,49 +0,0 @@
-{
-    "version": "0.2.0",
-    "configurations": [
-        {
-            "name": "Python: Attach",
-            "type": "python",
-            "request": "attach",
-            "port": 5678,  // or the port used by your Jupyter server
-            "justMyCode": false
-        },
-        {
-            "name": "Python: Aktuální soubor",
-            "type": "python",
-            "request": "launch",
-            "program": "${file}",
-            "cwd": "${workspaceFolder}",
-            "env": {
-                "PYTHONPATH": "${workspaceFolder}:${workspaceFolder}/bld"
-            },
-            "console": "integratedTerminal",
-            "justMyCode": false,
-            "python": "${command:python.interpreterPath}",
-            "internalConsoleOptions": "openOnSessionStart"
-        },
-        {
-            "name": "Python: Main",
-            "type": "python",
-            "request": "launch",
-            "program": "v2realbot/main.py",
-            "justMyCode": false
-        },
-        {
-            "name": "Python: File",
-            "type": "python",
-            "request": "launch",
-            "program": "${file}",
-            "justMyCode": false
-        },
-        {
-            "name": "Python: Attach",
-            "type": "python",
-            "request": "attach",
-            "connect": {
-                "host": "localhost",
-                "port": 5678
-            }
-        }
-    ]
-}

.vscode/settings.json

@@ -1,3 +0,0 @@
-{
-    "git.ignoreLimitWarning": true
-}

README.md

@@ -1,2 +1,3 @@
 # snippets
 A collection of reusable code snippets 
+* vbtpro [vbt-snippets.md](vbt-snippets.md)

deleteme.md

@@ -0,0 +1,4 @@
+[Fetching data](#Fetching%20data)
+
+# Fetching data
+fdfd

docs.md

@@ -0,0 +1,75 @@
+## MCP
+
+### Local vbtpro mcp server
+main code in `vectorbtpro.mcp_server` can run either in `stdio` or in `streamable-http mode`
+
+Run mcp server in streamable-http mode (currently manually in terminal as single process for all local mcp clients)
+```bash
+cd /Users/davidbrazda/Documents/Development/python/vectorbt.pro-2025.6.24
+source .venv/bin/activate 
+python -m vectorbtpro.mcp_server --transport streamable-http
+```
+
+Claude desktop (currently doesnt support remotes) can use it by `mcp-remote`
+
+```json
+      "vectorbtpro_server": {
+        "command": "npx",
+        "args": [
+          "-y", 
+          "mcp-remote", 
+          "http://localhost:8000/mcp"
+        ]
+      }
+```
+
+Other MCPs (Claude code, Gemini CLI)
+
+```json
+{
+  "mcpServers": {
+    "vectorbtpro_server": {
+      "transport": "http",
+      "url": "http://localhost:8000"
+    }
+  }
+}
+```
+
+
+## Claude Desktop
+
+**Settings:**
+
+ `/Users/davidbrazda/Library/Application Support/Claude/claude_desktop_config.json`
+ 
+**Logs:**
+
+ `/Users/davidbrazda/Library/Logs/Claude` (main.log,mcp.log, mcp_server-name.log)
+
+## Claude code
+[Doc](https://docs.anthropic.com/en/docs/claude-code/mcp)
+
+**Settings:**
+
+- User/global settings (`~/.claude.json`)
+- Project-specific MCP file (`.mcp.json`)
+
+
+**Dedicated MCP file:** ~/.claude/mcp_servers.json 
+
+**Logs:**
+
+## Claude Code Router
+[doc](https://github.com/musistudio/claude-code-router)
+
+**Settings:**
+
+To set models for each tasks
+
+`~/.claude-code-router/config.json`
+
+**Logs:**
+
+`~/.claude-code-router/claude-code-router.log`
+

features_targets.md

@@ -0,0 +1,259 @@
+Here goes the target features
+
+
+> [!NOTE] Poznámka
+> Contents
+
+
+> [!note]- poznám ka
+> neco
+> neco
+
+> [!example]- Graph: voaltility averagae slope
+> ![[Pasted image 20250630140635.png]]
+
+
+> [!example]- Graph: volatility average slope across 1d to 30d range of windows
+> ![[Volatility_average_slope.png]]
+
+# Things to try
+
+TODO:
+* lepsi labeling
+* continue here https://claude.ai/chat/b3ee78b6-9662-4f25-95f0-ecac4a78a41b
+* try model with other symbols
+* rey different retraining options (even hourly)
+
+Features:
+- add datetime features (useful for rush hour model)
+- add MT features as columns
+- use convolutional networks to create features (https://www.youtube.com/watch?v=6wK4q8QvsV4)
+Enhance model:
+* multi target see xgb doc
+* use SL with target price, with validy for few seconds
+* how handle imbalanced datase https://xgboost.readthedocs.io/en/stable/tutorials/param_tuning.html
+
+
+Target:
+- maybe add manual labeling
+
+# Features
+
+```python
+
+    def prepare_features(self, df: pd.DataFrame) -> tuple[pd.DataFrame, list]:
+        """Prepare enhanced features from input df with focus on predictive potential"""
+        features = pd.DataFrame(index=df.index)
+
+        # Original ohlcv added to features
+        features['close'] = df['close']
+        features['volume'] = df['volume']
+        features['trades_count'] = df['trades']
+        features['buy_volume'] = df['buyvolume']
+        features['sell_volume'] = df['sellvolume']
+        features['high'] = df['high']
+        features['low'] = df['low']
+        # features['log_return'] = np.log(features['close'] / features['close'].shift(1))
+        # features['returns_1'] = features['close'].pct_change()
+        # features['returns_5'] = features['close'].pct_change(5)
+        # features['returns_20'] = features['close'].pct_change(20)
+
+        def get_fib_windows():
+            """
+            #TODO based on real time (originally for 1s bars)
+
+            Generate Fibonacci sequence windows up to ~1 hour (3600 seconds)
+            Returns sequence: 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584
+            """
+            fib_windows = [3, 5]
+            while fib_windows[-1] < 3600/60:
+                next_fib = fib_windows[-1] + fib_windows[-2]
+                if next_fib > 3600/60:
+                    break
+                fib_windows.append(next_fib)
+            return fib_windows
+
+        fib_windows = get_fib_windows()
+        
+        # Base price and returns
+        features['log_return'] = np.log(features['close'] / features['close'].shift(1))
+        features['price_velocity'] = (features['close'] - features['close'].shift(1)) / 1.0  # per second
+        features['price_acceleration'] = features['price_velocity'] - features['price_velocity'].shift(1)
+        
+        # Fibonacci-based features
+        for window in fib_windows:
+            # Price features
+            features[f'log_return_{window}s'] = np.log(features['close'] / features['close'].shift(window))
+            features[f'volatility_{window}s'] = features['log_return'].rolling(window).std()
+            features[f'range_{window}s'] = (features['high'].rolling(window).max() - 
+                                        features['low'].rolling(window).min()) / features['close']
+            
+            # Volume features
+            features[f'volume_momentum_{window}s'] = (
+                features['volume'].rolling(window).mean() / 
+                features['volume'].rolling(window * 2).mean()
+            )
+            
+            features[f'buy_volume_momentum_{window}s'] = (
+                features['buy_volume'].rolling(window).mean() / 
+                features['buy_volume'].rolling(window * 2).mean()
+            )
+            
+            features[f'sell_volume_momentum_{window}s'] = (
+                features['sell_volume'].rolling(window).mean() / 
+                features['sell_volume'].rolling(window * 2).mean()
+            )
+            
+            # Trade features
+            features[f'trade_intensity_{window}s'] = (
+                features['trades_count'].rolling(window).mean() / 
+                features['trades_count'].rolling(window * 2).mean()
+            )
+            
+            features[f'avg_trade_size_{window}s'] = (
+                features['volume'].rolling(window).sum() / 
+                features['trades_count'].rolling(window).sum()
+            )
+            
+            # Order flow features
+            features[f'cum_volume_delta_{window}s'] = (
+                features['buy_volume'] - features['sell_volume']
+            ).rolling(window).sum()
+            
+            features[f'volume_pressure_{window}s'] = (
+                features['buy_volume'].rolling(window).sum() / 
+                features['sell_volume'].rolling(window).sum()
+            )
+            
+            # Price efficiency
+            features[f'price_efficiency_{window}s'] = (
+                np.abs(features['close'] - features['close'].shift(window)) /
+                (features['high'].rolling(window).max() - features['low'].rolling(window).min())
+            )
+            
+            # Moving averages and their crosses
+            features[f'sma_{window}s'] = features['close'].rolling(window).mean()
+            if window > 5:  # Create MA crosses with shorter timeframe
+                features[f'ma_cross_5_{window}s'] = (
+                    features['close'].rolling(5).mean() - 
+                    features['close'].rolling(window).mean()
+                )
+        
+        # MA-based features
+        ma_lengths = [5, 10, 20, 50]
+        for length in ma_lengths:
+            # Regular MAs
+            features[f'ma_{length}'] = features['close'].rolling(length).mean()
+            
+            # MA slopes (rate of change)
+            features[f'ma_{length}_slope'] = features[f'ma_{length}'].pct_change(3)
+            
+            # Price distance from MA
+            features[f'price_ma_{length}_dist'] = (features['close'] - features[f'ma_{length}']) / features[f'ma_{length}']
+            
+            # MA crossovers
+            if length > 5:
+                features[f'ma_5_{length}_cross'] = (features['ma_5'] - features[f'ma_{length}']) / features[f'ma_{length}']
+        
+        # MA convergence/divergence
+        features['ma_convergence'] = ((features['ma_5'] - features['ma_20']).abs() / 
+                                    features['ma_20'].rolling(10).mean())
+        
+        # Volatility features using MAs
+        features['ma_volatility'] = features['ma_5'].rolling(10).std() / features['ma_20']
+        
+        # MA momentum
+        features['ma_momentum'] = (features['ma_5'] / features['ma_5'].shift(5) - 1) * 100
+          
+        
+        # Cleanup and feature selection
+        features = features.replace([np.inf, -np.inf], np.nan)
+        
+        lookback = 1000
+        if len(features) > lookback:
+            rolling_corr = features.iloc[-lookback:].corr().abs()
+            upper = rolling_corr.where(np.triu(np.ones(rolling_corr.shape), k=1).astype(bool))
+            to_drop = [column for column in upper.columns if any(upper[column] > 0.95)]
+            print(f"Column highly correlated - maybe drop? {to_drop} ")
+            #features = features.drop(columns=to_drop)
+        
+        feature_columns = list(features.columns)
+        print(f"Features shape before dropna: {features.shape}")
+        
+        return features.dropna(), feature_columns
+```
+
+
+
+# Targets
+## Unbalanced classes
+
+```python
+from xgboost import XGBClassifier
+
+# Compute scale_pos_weight
+n_0 = sum(y_train == 0)
+n_1 = sum(y_train == 1)
+scale_pos_weight = n_0 / n_1
+
+model = XGBClassifier(scale_pos_weight=scale_pos_weight, ...)
+```
+
+
+```python
+    def create_target_regressor(self, df: pd.DataFrame) -> pd.Series:
+        """
+        https://claude.ai/chat/8e7fe81c-ddbe-4e64-9af0-2bc4764fc5f0
+
+        Creates enhanced target variable using adaptive returns based on market conditions.
+        Key improvements:
+        1. Multi-timeframe momentum approach
+        2. Volume-volatility regime adaptation
+        3. Trend-following vs mean-reversion regime detection
+        4. Noise reduction through sophisticated filtering
+        
+        Parameters:
+        -----------
+        df : pd.DataFrame
+            Features df containing required columns: 'close', 'volume', volatility features
+        
+        Returns:
+        --------
+        pd.Series
+            Enhanced target variable with cross-day targets removed
+        """
+
+        future_bars= self.config.forward_bars
+
+        future_ma_fast = df['close'].shift(-future_bars).rolling(5).mean()
+        
+        # Calculate forward returns (original approach)
+        forward_returns = df['close'].shift(-future_bars) / df['close'] - 1
+        
+        target =  forward_returns
+
+       # 6. Advanced noise reduction
+        # Use exponential moving standard deviation for dynamic thresholds
+        target_std = target.ewm(span=50, min_periods=20).std()
+        
+        # Adaptive thresholds based on rolling standard deviation
+        upper_clip = 2.5 * target_std
+        lower_clip = -2.5 * target_std
+        
+        # Apply soft clipping using hyperbolic tangent
+        target = target_std * np.tanh(target / target_std)
+        
+        # Final hard clips for extreme outliers
+        target = target.clip(lower=lower_clip, upper=upper_clip)
+
+
+        # 7. Remove cross-day targets and intraday seasonality
+        target = self.remove_crossday_targets(target, df, future_bars)
+
+        #only 10% of extreme values from both sides are kept
+        #target = target.where((target > target.quantile(0.9)) | (target < target.quantile(0.1)), 0)
+
+        print("after target generation", target.index[[0, -1]])
+        
+        return target
+```

ml-snippets.md

@@ -0,0 +1,317 @@
+- [Features](#features)
+  - [Features analysis](#features-analysis)
+    - [Target to classes](#target-to-classes)
+  - [Features importance](#features-importance)
+  - [Features selection](#features-selection)
+- [Prediction](#prediction)
+  - [evaluation](#evaluation)
+    - [calculated returns based on various probability prediction thresholda](#calculated-returns-based-on-various-probability-prediction-thresholda)
+    - [cumulative returns bases od prob predictions](#cumulative-returns-bases-od-prob-predictions)
+  - [charts](#charts)
+
+
+
+
+
+# Features
+
+## Features analysis
+
+```python
+# Calculate different percentiles
+percentiles = [1, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 95, 99]
+print("\nPercentiles:")
+for p in percentiles:
+    print(f"{p}th percentile: {df['target'].quantile(p/100):.6f}")
+
+# Plot distribution
+plt.figure(figsize=(15, 10))
+
+# Plot 1: Overall distribution
+plt.subplot(2, 2, 1)
+sns.histplot(df['target'], bins=100)
+plt.title('Distribution of Returns')
+plt.axvline(x=0, color='r', linestyle='--', alpha=0.5)
+
+# Plot 2: Distribution with potential thresholds
+plt.subplot(2, 2, 2)
+sns.histplot(df['target'], bins=100)
+plt.title('Distribution with Potential Thresholds')
+
+# Add lines for different standard deviations
+std = df['target'].std()
+mean = df['target'].mean()
+for i in [0.5, 1.0, 1.5]:
+    plt.axvline(x=mean + i*std, color='g', linestyle='--', alpha=0.3, label=f'+{i} std')
+    plt.axvline(x=mean - i*std, color='r', linestyle='--', alpha=0.3, label=f'-{i} std')
+plt.legend()
+
+# Let's try different threshold approaches
+# Approach 1: Standard deviation based
+std_multiplier = 0.2
+std_threshold = std_multiplier * std
+labels_std = np.where(df['target'] > std_threshold, 1,
+                     np.where(df['target'] < -std_threshold, -1, 0))
+
+# Approach 2: Percentile based
+percentile_threshold = 0.2  # top/bottom 20%
+top_threshold = df['target'].quantile(1 - percentile_threshold)
+bottom_threshold = df['target'].quantile(percentile_threshold)
+labels_percentile = np.where(df['target'] > top_threshold, 1,
+                           np.where(df['target'] < bottom_threshold, -1, 0))
+
+# Plot 3: Distribution of STD-based classes
+plt.subplot(2, 2, 3)
+sns.histplot(data=pd.DataFrame({'return': df['target'], 'class': labels_std}), 
+            x='return', hue='class', bins=100)
+plt.title(f'Classes Based on {std_multiplier} Standard Deviation')
+plt.axvline(x=std_threshold, color='g', linestyle='--', alpha=0.5)
+plt.axvline(x=-std_threshold, color='r', linestyle='--', alpha=0.5)
+
+# Plot 4: Distribution of Percentile-based classes
+plt.subplot(2, 2, 4)
+sns.histplot(data=pd.DataFrame({'return': df['target'], 'class': labels_percentile}), 
+            x='return', hue='class', bins=100)
+plt.title(f'Classes Based on {percentile_threshold*100}th Percentiles')
+plt.axvline(x=top_threshold, color='g', linestyle='--', alpha=0.5)
+plt.axvline(x=bottom_threshold, color='r', linestyle='--', alpha=0.5)
+
+plt.tight_layout()
+plt.show()
+
+# Print class distributions
+print("\nClass Distribution (STD-based):")
+print(pd.Series(labels_std).value_counts(normalize=True))
+
+print("\nClass Distribution (Percentile-based):")
+print(pd.Series(labels_percentile).value_counts(normalize=True))
+
+# Calculate mean return for each class
+print("\nMean Return by Class (STD-based):")
+std_df = pd.DataFrame({'return': df['target'], 'class': labels_std})
+print(std_df.groupby('class')['return'].mean())
+
+print("\nMean Return by Class (Percentile-based):")
+perc_df = pd.DataFrame({'return': df['target'], 'class': labels_percentile})
+print(perc_df.groupby('class')['return'].mean())
+```
+<img src="image-1.png" alt="Target distributions" width="300"/>
+
+### Target to classes
+
+Based on std dev
+
+```python
+# Read and prepare the data
+df = pd.read_csv('model_data.csv')
+df = df.drop('ts_event', axis=1)
+
+# Separate features and target
+X = df.drop('target', axis=1)
+y = df['target']
+
+# Split the data first so we only use train data statistics for thresholds
+X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+# Calculate threshold based on training data only
+train_std = y_train.std()
+threshold = 0.2 * train_std
+
+# Transform targets into classes (update this function) instead of -1,0,1 do 0,1,2
+def create_labels(y, threshold):
+    return np.where(y > threshold, 2,
+                   np.where(y < -threshold, 0, 1))
+
+y_train_classes = create_labels(y_train, threshold)
+y_test_classes = create_labels(y_test, threshold)
+
+# Print class distribution
+print("Training Class Distribution:")
+print(pd.Series(y_train_classes).value_counts(normalize=True))
+print("\nTest Class Distribution:")
+print(pd.Series(y_test_classes).value_counts(normalize=True))
+```
+
+based on percentile/threshold
+
+## Features importance
+
+```python
+#XGB top 20 feature importance
+feature_importance = pd.DataFrame({
+    'feature': X.columns,
+    'importance': xgb_model.feature_importances_
+})
+feature_importance = feature_importance.sort_values('importance', ascending=False).head(20)
+
+plt.figure(figsize=(12, 6))
+sns.barplot(x='importance', y='feature', data=feature_importance)
+plt.title('Top 20 Most Important Features')
+plt.xlabel('Feature Importance')
+plt.tight_layout()
+plt.show()
+
+```
+
+## Features selection
+
+# Prediction
+
+## evaluation
+
+```python
+# Calculate directional accuracy
+directional_accuracy = (np.sign(y_pred) == np.sign(y_test)).mean()
+print(f"Directional Accuracy: {directional_accuracy:.4f}")
+
+#confusion matrix
+from sklearn.metrics import confusion_matrix
+# Plot confusion matrix
+plt.figure(figsize=(10, 8))
+cm = confusion_matrix(y_test_classes, y_pred)
+sns.heatmap(cm, annot=True, fmt='d', cmap='Blues')
+plt.title('Confusion Matrix')
+plt.ylabel('True Label')
+plt.xlabel('Predicted Label')
+plt.show()
+```
+### calculated returns based on various probability prediction thresholda
+
+```python
+# .predict_proba() gives the probabilities for each class
+print("Predicted probabilities:", model.predict_proba(X_test))
+# Output example:
+# [
+#  [0.35, 0.65],  # 35% not spam, 65% spam
+#  [0.70, 0.30],  # 70% not spam, 30% spam
+#  [0.45, 0.55],  # 45% not spam, 55% spam
+# ]
+```
+
+Chart probabilities
+
+```python
+# Predict probabilities for each class
+probabilities = model.predict_proba(X_test)  # Shape: (n_samples, n_classes)
+results_df = pd.DataFrame({
+    'Date': dates_test,
+    'Short Probability': probabilities[:, 0],   # Probability of class 0 (short)
+    'Neutral Probability': probabilities[:, 1], # Probability of class 1 (neutral)
+    'Long Probability': probabilities[:, 2]     # Probability of class 2 (long)
+}).sort_values(by='Date')  # Sort by date for time series plotting
+
+fig = go.Figure()
+
+# Add lines for each class probability
+fig.add_trace(go.Scatter(
+    x=results_df['Date'], y=results_df['Short Probability'],
+    mode='lines', name='Short (Class 0)', line=dict(color='red')
+))
+
+fig.add_trace(go.Scatter(
+    x=results_df['Date'], y=results_df['Neutral Probability'],
+    mode='lines', name='Neutral (Class 1)', line=dict(color='orange')
+))
+
+fig.add_trace(go.Scatter(
+    x=results_df['Date'], y=results_df['Long Probability'],
+    mode='lines', name='Long (Class 2)', line=dict(color='green')
+))
+
+# Add title and labels
+fig.update_layout(
+    title="Time Series of Predicted Class Probabilities",
+    xaxis_title="Date",
+    yaxis_title="Probability",
+    legend_title="Class"
+)
+
+fig.show()
+
+```
+
+### cumulative returns bases od prob predictions
+```python
+# Calculate returns based on probablity predictions
+def calculate_returns(predictions, actual_returns, confidence_threshold=0.0):
+    pred_probs = final_model.predict_proba(X_test_selected)
+    max_probs = np.max(pred_probs, axis=1)
+    
+    # Only take positions when confidence exceeds threshold
+    positions = np.zeros_like(predictions, dtype=float)
+    confident_mask = max_probs > confidence_threshold
+    
+    # Convert predictions 0->-1, 2->1 for returns calculation
+    adj_predictions = np.where(predictions == 2, 1, np.where(predictions == 0, -1, 0))
+    positions[confident_mask] = adj_predictions[confident_mask]
+    
+    returns = positions * actual_returns
+    return returns, np.mean(confident_mask)
+
+# Test different confidence thresholds
+confidence_thresholds = [0.4, 0.5, 0.6, 0.7, 0.8]
+results = []
+
+for conf_threshold in confidence_thresholds:
+    returns, coverage = calculate_returns(y_pred, y_test.values, conf_threshold)
+    
+    # Calculate metrics
+    sharpe = np.sqrt(252) * returns.mean() / returns.std()
+    accuracy = accuracy_score(y_test_classes[returns != 0], 
+                            y_pred[returns != 0])
+    
+    results.append({
+        'confidence_threshold': conf_threshold,
+        'sharpe': sharpe,
+        'accuracy': accuracy,
+        'coverage': coverage
+    })
+
+##Plot difference confidence threshodls
+# Plot cumulative returns
+plt.figure(figsize=(12, 6))
+for th in confidence_thresholds:
+    returns, _ = calculate_returns(y_pred, y_test.values, th)  # Using 0.6 confidence threshold
+    cumulative_returns = (1 + returns).cumprod()
+    plt.plot(cumulative_returns)
+plt.title('Cumulative Returns (0.6 confidence threshold)')
+plt.xlabel('Trade Number')
+plt.ylabel('Cumulative Return')
+plt.grid(True)
+plt.show()
+
+
+results_df = pd.DataFrame(results)
+print("\nPerformance at different confidence thresholds:")
+print(results_df)
+
+# Plot feature importance
+importance_df = pd.DataFrame({
+    'feature': selected_features,
+    'importance': final_model.feature_importances_
+})
+importance_df = importance_df.sort_values('importance', ascending=False)
+
+plt.figure(figsize=(12, 6))
+sns.barplot(x='importance', y='feature', data=importance_df)
+plt.title('Feature Importance')
+plt.xlabel('Importance')
+plt.tight_layout()
+plt.show()
+
+```
+
+
+## charts
+
+```python
+# Actual vs predicted values
+plt.figure(figsize=(10, 6))
+plt.scatter(y_test, y_pred, alpha=0.5)
+plt.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--', lw=2)
+plt.xlabel('Actual Returns')
+plt.ylabel('Predicted Returns')
+plt.title('Actual vs Predicted Returns')
+plt.tight_layout()
+plt.show()
+```

pandas-snippets.md

@@ -0,0 +1,9 @@
+- [Basics](#basics)
+
+# Basics
+
+```python
+df.loc[:, "D"] = np.array([5] * len(df))
+df.loc[:, ["A", "B"]]
+df.sort_index(axis=1, ascending=False) #axis=1 sort columns by labels, axis=0 sort by index
+```

setup.py

@@ -1,7 +1,7 @@
 from setuptools import find_packages, setup
 
 setup(name='snippets',
-      version='0.1',
+      version='0.2.2',
       description='Snippets for vbtpro',
       author='David Brazda',
       author_email='davidbrazda61@gmail.com',

vbt-topics/vbt-custom-metrics-analysis.md

@@ -0,0 +1,893 @@
+Custom metrics in VectorBT PRO, which is a powerful feature for extending portfolio analysis beyond the built-in metrics.Now let me provide a comprehensive elaboration on custom metrics in VectorBT PRO:
+
+# Custom Metrics in VectorBT PRO
+
+Custom metrics in VectorBT PRO provide a powerful way to extend portfolio analysis beyond the built-in metrics. They allow you to calculate domain-specific metrics, implement proprietary performance measures, or create metrics tailored to your specific trading strategies.
+
+## 1. Understanding the Metrics System
+
+### Built-in Metrics Structure
+VectorBT PRO uses a configuration-based approach where metrics are stored in `Portfolio.metrics` as a `HybridConfig`:
+
+```python
+# View all available metrics
+print(vbt.Portfolio.metrics)
+
+# Get specific metric configuration
+print(vbt.Portfolio.metrics['sharpe_ratio'])
+```
+
+### Metric Configuration Structure
+Each metric is defined as a dictionary with specific keys:
+
+```python
+metric_config = {
+    'title': 'My Custom Metric',        # Display name
+    'calc_func': calculation_function,   # Function to calculate the metric
+    'tags': ['custom', 'risk'],         # Tags for filtering
+    'apply_to_timedelta': False,        # Whether to convert to timedelta
+    'agg_func': None,                   # Aggregation function
+    'resolve_calc_func': True,          # Whether to resolve attributes
+    # ... other configuration options
+}
+```
+
+## 2. Creating Custom Metrics
+
+### Method 1: Simple Function-Based Metrics
+
+```python
+# Add a simple custom metric
+vbt.Portfolio.metrics['total_bars'] = dict(
+    title='Total Bars',
+    calc_func=lambda self: len(self.wrapper.index)
+)
+
+# Add skewness and kurtosis
+vbt.Portfolio.metrics['skew'] = dict(
+    title='Skew', 
+    calc_func='returns.skew'
+)
+vbt.Portfolio.metrics['kurtosis'] = dict(
+    title='Kurtosis', 
+    calc_func='returns.kurtosis'
+)
+```
+
+### Method 2: Complex Custom Calculations
+
+```python
+# Custom metric with multiple parameters
+def total_return_no_fees(self, orders):
+    """Calculate total return without fees"""
+    return (self.total_profit + orders.fees.sum()) / self.get_init_cash() * 100
+
+vbt.Portfolio.metrics['total_return_no_fees'] = dict(
+    title='Total Return (No Fees) [%]',
+    calc_func=total_return_no_fees,
+    resolve_orders=True  # Automatically resolve orders parameter
+)
+```
+
+### Method 3: Using Lambda Functions with Settings
+
+```python
+# PnL in dollar terms (for futures trading)
+vbt.Portfolio.metrics['pnl_dollars'] = dict(
+    title='PnL ($)',
+    calc_func=lambda self, settings: (self.value.iloc[-1] - self.value.iloc[0]) * 50,
+    resolve_calc_func=False  # Don't resolve attributes automatically
+)
+```
+
+## 3. Advanced Custom Metrics
+
+### Quantile-Based Metrics
+```python
+def value_at_risk_custom(returns, confidence_level=0.05):
+    """Custom VaR calculation"""
+    return returns.quantile(confidence_level)
+
+vbt.Portfolio.metrics['custom_var'] = dict(
+    title='Custom VaR (5%)',
+    calc_func=value_at_risk_custom,
+    resolve_returns=True,
+    confidence_level=0.05
+)
+```
+
+### Multi-Component Metrics
+```python
+def comprehensive_trade_stats(trades):
+    """Return multiple trade statistics"""
+    return {
+        'long_trades': trades.direction_long.count(),
+        'short_trades': trades.direction_short.count(),
+        'long_pnl': trades.direction_long.pnl.sum(),
+        'short_pnl': trades.direction_short.pnl.sum(),
+        'avg_trade_duration': trades.duration.mean()
+    }
+
+vbt.Portfolio.metrics['trade_breakdown'] = dict(
+    title='Trade Breakdown',
+    calc_func=comprehensive_trade_stats,
+    resolve_trades=True
+)
+```
+
+### Time-Based Metrics
+```python
+def monthly_returns_volatility(returns):
+    """Calculate monthly returns volatility"""
+    monthly_returns = returns.resample('M').sum()
+    return monthly_returns.std() * np.sqrt(12)
+
+vbt.Portfolio.metrics['monthly_vol'] = dict(
+    title='Monthly Volatility',
+    calc_func=monthly_returns_volatility,
+    resolve_returns=True
+)
+```
+
+## 4. Metric Resolution and Parameters
+
+### Automatic Parameter Resolution
+VectorBT PRO can automatically resolve portfolio attributes as parameters:
+
+```python
+# These parameters will be automatically resolved:
+vbt.Portfolio.metrics['custom_metric'] = dict(
+    title='Custom Metric',
+    calc_func=lambda returns, trades, orders: calculation_logic(returns, trades, orders),
+    resolve_returns=True,    # Passes self.returns
+    resolve_trades=True,     # Passes self.trades
+    resolve_orders=True      # Passes self.orders
+)
+```
+
+### Common Resolvable Parameters
+- `self` - The portfolio instance
+- `returns` - Portfolio returns
+- `trades` - Trade records
+- `orders` - Order records
+- `drawdowns` - Drawdown records
+- `value` - Portfolio value
+- `close` - Close prices
+- `init_cash` - Initial cash
+- `total_profit` - Total profit
+- `wrapper` - Array wrapper (for index/column info)
+
+## 5. Global vs Instance-Level Metrics
+
+### Global Metrics (Class-Level)
+```python
+# Add to all future Portfolio instances
+vbt.Portfolio.metrics['my_metric'] = metric_config
+
+# Or modify settings globally
+vbt.settings.portfolio['stats']['metrics'] = list(vbt.Portfolio.metrics.items()) + [
+    ('my_metric', metric_config)
+]
+```
+
+### Instance-Level Metrics
+```python
+# Add to specific portfolio instance
+pf._metrics['my_metric'] = metric_config
+
+# Then use it
+pf.stats(['my_metric'])
+```
+
+## 6. Using Custom Metrics
+
+### Basic Usage
+```python
+# Calculate specific custom metrics
+pf.stats(['total_bars', 'skew', 'kurtosis'])
+
+# Calculate all metrics including custom ones
+pf.stats('all')
+
+# Filter by tags
+pf.stats(tags=['custom'])
+```
+
+### Advanced Usage with Settings
+```python
+# Use custom metrics in optimization
+results = []
+for param in parameter_combinations:
+    pf = vbt.Portfolio.from_signals(close, entries, exits, **param)
+    stats = pf.stats(['total_return', 'sharpe_ratio', 'my_custom_metric'])
+    results.append(stats)
+
+# Create comparison DataFrame
+comparison_df = pd.DataFrame(results)
+```
+
+## 7. Real-World Examples
+
+### Futures Trading Metrics
+```python
+# Point-based P&L for futures
+vbt.Portfolio.metrics['pnl_points'] = dict(
+    title='P&L (Points)',
+    calc_func=lambda self: (self.value.iloc[-1] - self.value.iloc[0]) / self.close.iloc[0] * 10000
+)
+
+# Risk-adjusted return for futures
+vbt.Portfolio.metrics['risk_adjusted_return'] = dict(
+    title='Risk Adjusted Return',
+    calc_func=lambda self, returns: self.total_return / returns.std() * np.sqrt(252),
+    resolve_returns=True
+)
+```
+
+### Intraday Strategy Metrics
+```python
+# Time-of-day analysis
+def intraday_performance(orders):
+    """Analyze performance by hour of day"""
+    order_df = orders.records_readable
+    order_df['hour'] = order_df.index.hour
+    return order_df.groupby('hour')['PnL'].mean()
+
+vbt.Portfolio.metrics['hourly_performance'] = dict(
+    title='Hourly Performance',
+    calc_func=intraday_performance,
+    resolve_orders=True
+)
+```
+
+### Market Regime Metrics
+```python
+def regime_performance(returns, benchmark_returns):
+    """Performance in different market regimes"""
+    bull_mask = benchmark_returns > benchmark_returns.quantile(0.6)
+    bear_mask = benchmark_returns < benchmark_returns.quantile(0.4)
+    
+    return {
+        'bull_return': returns[bull_mask].mean(),
+        'bear_return': returns[bear_mask].mean(),
+        'bull_sharpe': returns[bull_mask].mean() / returns[bull_mask].std() * np.sqrt(252),
+        'bear_sharpe': returns[bear_mask].mean() / returns[bear_mask].std() * np.sqrt(252)
+    }
+
+vbt.Portfolio.metrics['regime_analysis'] = dict(
+    title='Market Regime Analysis',
+    calc_func=regime_performance,
+    resolve_returns=True,
+    resolve_bm_returns=True
+)
+```
+
+## 8. Best Practices
+
+### 1. Naming Conventions
+- Use descriptive names: `monthly_volatility` instead of `mv`
+- Include units in title: `'Max Drawdown [%]'`
+- Use consistent naming patterns
+
+### 2. Error Handling
+```python
+def robust_metric(returns):
+    """Metric with error handling"""
+    try:
+        if len(returns) < 2:
+            return np.nan
+        return returns.std() * np.sqrt(252)
+    except Exception as e:
+        print(f"Error calculating metric: {e}")
+        return np.nan
+```
+
+### 3. Performance Optimization
+```python
+# Use vectorized operations
+def efficient_metric(returns):
+    """Efficient vectorized calculation"""
+    return returns.rolling(30).std().mean()
+
+# Avoid loops when possible
+def inefficient_metric(returns):
+    """Avoid this approach"""
+    results = []
+    for i in range(len(returns)):
+        results.append(some_calculation(returns.iloc[i]))
+    return np.mean(results)
+```
+
+### 4. Documentation
+```python
+vbt.Portfolio.metrics['documented_metric'] = dict(
+    title='Well Documented Metric',
+    calc_func=lambda returns: returns.std() * np.sqrt(252),
+    resolve_returns=True,
+    tags=['custom', 'risk', 'volatility'],
+    # Add description in comments or docstrings
+)
+```
+
+## 9. Common Pitfalls and Solutions
+
+### Pitfall 1: Metric Not Available After Creation
+```python
+# ❌ Wrong: Metric added after portfolio creation
+pf = vbt.Portfolio.from_signals(...)
+vbt.Portfolio.metrics['my_metric'] = metric_config
+pf.stats(['my_metric'])  # KeyError!
+
+# ✅ Correct: Add metric before portfolio creation
+vbt.Portfolio.metrics['my_metric'] = metric_config
+pf = vbt.Portfolio.from_signals(...)
+pf.stats(['my_metric'])  # Works!
+```
+
+### Pitfall 2: Incorrect Parameter Resolution
+```python
+# ❌ Wrong: Using external variables
+portfolio_instance = some_portfolio
+vbt.Portfolio.metrics['bad_metric'] = dict(
+    calc_func=lambda self: portfolio_instance.total_return  # External reference
+)
+
+# ✅ Correct: Using self parameter
+vbt.Portfolio.metrics['good_metric'] = dict(
+    calc_func=lambda self: self.total_return  # Self reference
+)
+```
+
+### Pitfall 3: Missing Error Handling
+```python
+# ❌ Wrong: No error handling
+def risky_metric(trades):
+    return trades.pnl.sum() / trades.duration.mean()  # Division by zero possible
+
+# ✅ Correct: With error handling
+def safe_metric(trades):
+    if len(trades) == 0 or trades.duration.mean() == 0:
+        return np.nan
+    return trades.pnl.sum() / trades.duration.mean()
+```
+
+Custom metrics in VectorBT PRO provide unlimited flexibility to analyze your trading strategies exactly how you need. They integrate seamlessly with the existing stats system and can be used in optimization, comparison, and reporting workflows.
+
+I'll provide a comprehensive analysis of VectorBT PRO's `pf.trades` analysis capabilities, with a focus on the specific metrics you mentioned.# Comprehensive VectorBT PRO `pf.trades` Analysis
+
+The `pf.trades` object in VectorBT PRO provides extensive capabilities for analyzing trading performance. Here's a comprehensive guide focusing on directional analysis, temporal patterns, and advanced trade analytics.
+
+## 1. Basic Trade Analysis
+
+### Trade Counts by Direction
+```python
+# Basic trade counts
+total_trades = pf.trades.count()
+long_trades = pf.trades.direction_long.count()
+short_trades = pf.trades.direction_short.count()
+
+print(f"Total trades: {total_trades}")
+print(f"Long trades: {long_trades}")
+print(f"Short trades: {short_trades}")
+
+# Alternative using records
+trade_records = pf.trades.records_readable
+direction_counts = trade_records['Direction'].value_counts()
+print(f"\nDirection breakdown:\n{direction_counts}")
+```
+
+### P&L Analysis by Direction
+```python
+# Total P&L by direction
+long_pnl = pf.trades.direction_long.pnl.sum()
+short_pnl = pf.trades.direction_short.pnl.sum()
+total_pnl = pf.trades.pnl.sum()
+
+print(f"Long P&L: {long_pnl:.2f}")
+print(f"Short P&L: {short_pnl:.2f}")
+print(f"Total P&L: {total_pnl:.2f}")
+
+# P&L statistics by direction
+long_stats = pf.trades.direction_long.pnl.describe()
+short_stats = pf.trades.direction_short.pnl.describe()
+```
+
+## 2. Daily P&L Analysis
+
+### Daily P&L Calculation
+```python
+# Method 1: Using trade records with date grouping
+trade_records = pf.trades.records_readable
+trade_records['exit_date'] = trade_records.index.date
+
+# Daily P&L overall
+daily_pnl = trade_records.groupby('exit_date')['PnL'].sum()
+
+# Daily P&L by direction
+daily_pnl_by_direction = trade_records.groupby(['exit_date', 'Direction'])['PnL'].sum().unstack(fill_value=0)
+
+print("Daily P&L by Direction:")
+print(daily_pnl_by_direction.head())
+```
+
+### Daily P&L for Each Direction
+```python
+# Separate long and short daily P&L
+long_trades_records = trade_records[trade_records['Direction'] == 'Long']
+short_trades_records = trade_records[trade_records['Direction'] == 'Short']
+
+daily_long_pnl = long_trades_records.groupby('exit_date')['PnL'].sum()
+daily_short_pnl = short_trades_records.groupby('exit_date')['PnL'].sum()
+
+# Combine into comprehensive daily analysis
+daily_analysis = pd.DataFrame({
+    'Total_PnL': daily_pnl,
+    'Long_PnL': daily_long_pnl,
+    'Short_PnL': daily_short_pnl,
+    'Long_Trades': long_trades_records.groupby('exit_date').size(),
+    'Short_Trades': short_trades_records.groupby('exit_date').size()
+}).fillna(0)
+
+print("Daily Trade Analysis:")
+print(daily_analysis.head())
+```
+
+## 3. Hourly P&L Analysis by Exit Time
+
+### Hourly P&L by Direction
+```python
+# Extract hour from exit time
+trade_records = pf.trades.records_readable
+trade_records['exit_hour'] = trade_records.index.hour
+
+# Hourly P&L analysis
+hourly_pnl_analysis = trade_records.groupby(['exit_hour', 'Direction']).agg({
+    'PnL': ['sum', 'mean', 'count'],
+    'Return': ['mean', 'std']
+}).round(4)
+
+print("Hourly P&L Analysis by Direction:")
+print(hourly_pnl_analysis)
+
+# Separate analysis for each direction
+hourly_long_pnl = trade_records[trade_records['Direction'] == 'Long'].groupby('exit_hour')['PnL'].agg(['sum', 'mean', 'count'])
+hourly_short_pnl = trade_records[trade_records['Direction'] == 'Short'].groupby('exit_hour')['PnL'].agg(['sum', 'mean', 'count'])
+
+print("\nHourly Long P&L:")
+print(hourly_long_pnl)
+
+print("\nHourly Short P&L:")
+print(hourly_short_pnl)
+```
+
+### Advanced Hourly Analysis
+```python
+# Create comprehensive hourly performance matrix
+def hourly_performance_analysis(trades_records):
+    """Comprehensive hourly performance analysis"""
+    
+    # Add time components
+    trades_records['exit_hour'] = trades_records.index.hour
+    trades_records['entry_hour'] = pd.to_datetime(trades_records['Entry Index']).dt.hour
+    
+    # Hourly exit analysis
+    hourly_stats = trades_records.groupby(['exit_hour', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count', 'std'],
+        'Return': ['mean', 'std'],
+        'Size': 'mean'
+    }).round(4)
+    
+    return hourly_stats
+
+hourly_performance = hourly_performance_analysis(trade_records)
+```
+
+## 4. Day of Week Analysis
+
+### P&L by Day of Week and Direction
+```python
+# Add day of week analysis
+trade_records['exit_day_of_week'] = trade_records.index.day_name()
+trade_records['exit_weekday'] = trade_records.index.weekday  # 0=Monday, 6=Sunday
+
+# Day of week P&L analysis
+dow_analysis = trade_records.groupby(['exit_day_of_week', 'Direction']).agg({
+    'PnL': ['sum', 'mean', 'count'],
+    'Return': ['mean', 'std'],
+    'Size': 'mean'
+}).round(4)
+
+print("Day of Week Analysis:")
+print(dow_analysis)
+
+# Pivot for easier viewing
+dow_pivot = trade_records.pivot_table(
+    index='exit_day_of_week',
+    columns='Direction',
+    values='PnL',
+    aggfunc=['sum', 'mean', 'count'],
+    fill_value=0
+)
+
+print("\nDay of Week Pivot Analysis:")
+print(dow_pivot)
+```
+
+### Advanced Day of Week Patterns
+```python
+# Create comprehensive day of week analysis
+def day_of_week_analysis(trades_records):
+    """Comprehensive day of week performance analysis"""
+    
+    # Ensure we have day names in proper order
+    day_order = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
+    
+    trades_records['exit_day_name'] = trades_records.index.day_name()
+    
+    # Group by day and direction
+    dow_stats = trades_records.groupby(['exit_day_name', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count', 'std'],
+        'Return': ['mean', 'std'],
+        'Size': 'mean',
+        'Entry Fees': 'mean',
+        'Exit Fees': 'mean'
+    }).round(4)
+    
+    # Reorder by day
+    dow_stats = dow_stats.reindex(day_order, level=0)
+    
+    return dow_stats
+
+dow_comprehensive = day_of_week_analysis(trade_records)
+```
+
+## 5. Advanced Temporal Analysis
+
+### Combined Time Pattern Analysis
+```python
+# Create comprehensive time pattern analysis
+def comprehensive_time_analysis(pf):
+    """Complete temporal analysis of trades"""
+    
+    trades_records = pf.trades.records_readable
+    
+    # Add all time components
+    trades_records['exit_hour'] = trades_records.index.hour
+    trades_records['exit_day_name'] = trades_records.index.day_name()
+    trades_records['exit_month'] = trades_records.index.month
+    trades_records['exit_date'] = trades_records.index.date
+    
+    # 1. Hourly analysis
+    hourly_stats = trades_records.groupby(['exit_hour', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count'],
+        'Return': ['mean', 'std']
+    }).round(4)
+    
+    # 2. Daily analysis
+    daily_stats = trades_records.groupby(['exit_day_name', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count'],
+        'Return': ['mean', 'std']
+    }).round(4)
+    
+    # 3. Monthly analysis
+    monthly_stats = trades_records.groupby(['exit_month', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count'],
+        'Return': ['mean', 'std']
+    }).round(4)
+    
+    # 4. Combined hour-day analysis
+    hour_day_stats = trades_records.groupby(['exit_day_name', 'exit_hour', 'Direction']).agg({
+        'PnL': ['sum', 'mean', 'count']
+    }).round(4)
+    
+    return {
+        'hourly': hourly_stats,
+        'daily': daily_stats,
+        'monthly': monthly_stats,
+        'hour_day': hour_day_stats
+    }
+
+# Execute comprehensive analysis
+time_analysis = comprehensive_time_analysis(pf)
+
+# Display results
+print("=== HOURLY ANALYSIS ===")
+print(time_analysis['hourly'])
+print("\n=== DAILY ANALYSIS ===")
+print(time_analysis['daily'])
+print("\n=== MONTHLY ANALYSIS ===")
+print(time_analysis['monthly'])
+```
+
+## 6. Custom Metrics for Trade Analysis
+
+### Custom Trade Metrics
+```python
+# Add custom metrics to Portfolio for directional analysis
+vbt.Portfolio.metrics['long_trade_count'] = dict(
+    title='Long Trade Count',
+    calc_func=lambda trades: trades.direction_long.count(),
+    resolve_trades=True
+)
+
+vbt.Portfolio.metrics['short_trade_count'] = dict(
+    title='Short Trade Count',
+    calc_func=lambda trades: trades.direction_short.count(),
+    resolve_trades=True
+)
+
+vbt.Portfolio.metrics['long_pnl_total'] = dict(
+    title='Long P&L Total',
+    calc_func=lambda trades: trades.direction_long.pnl.sum(),
+    resolve_trades=True
+)
+
+vbt.Portfolio.metrics['short_pnl_total'] = dict(
+    title='Short P&L Total',
+    calc_func=lambda trades: trades.direction_short.pnl.sum(),
+    resolve_trades=True
+)
+
+# Temporal metrics
+vbt.Portfolio.metrics['best_hour_pnl'] = dict(
+    title='Best Hour P&L',
+    calc_func=lambda trades: trades.records_readable.groupby(trades.records_readable.index.hour)['PnL'].sum().max(),
+    resolve_trades=True
+)
+
+vbt.Portfolio.metrics['worst_hour_pnl'] = dict(
+    title='Worst Hour P&L',
+    calc_func=lambda trades: trades.records_readable.groupby(trades.records_readable.index.hour)['PnL'].sum().min(),
+    resolve_trades=True
+)
+```
+
+## 7. Performance Analysis Functions
+
+### Comprehensive Trade Performance Function
+```python
+def analyze_trade_performance(pf):
+    """Comprehensive trade performance analysis"""
+    
+    trades = pf.trades
+    records = trades.records_readable
+    
+    # Basic directional statistics
+    direction_stats = {
+        'Long': {
+            'count': trades.direction_long.count(),
+            'total_pnl': trades.direction_long.pnl.sum(),
+            'avg_pnl': trades.direction_long.pnl.mean(),
+            'win_rate': trades.direction_long.win_rate,
+            'profit_factor': trades.direction_long.profit_factor
+        },
+        'Short': {
+            'count': trades.direction_short.count(),
+            'total_pnl': trades.direction_short.pnl.sum(),
+            'avg_pnl': trades.direction_short.pnl.mean(),
+            'win_rate': trades.direction_short.win_rate,
+            'profit_factor': trades.direction_short.profit_factor
+        }
+    }
+    
+    # Temporal analysis
+    records['exit_hour'] = records.index.hour
+    records['exit_day'] = records.index.day_name()
+    records['exit_date'] = records.index.date
+    
+    # Hourly P&L by direction
+    hourly_pnl = records.groupby(['exit_hour', 'Direction'])['PnL'].agg(['sum', 'mean', 'count'])
+    
+    # Daily P&L by direction
+    daily_pnl = records.groupby(['exit_day', 'Direction'])['PnL'].agg(['sum', 'mean', 'count'])
+    
+    # Date-based P&L
+    date_pnl = records.groupby(['exit_date', 'Direction'])['PnL'].agg(['sum', 'mean', 'count'])
+    
+    # Best/worst performing times
+    best_hours = records.groupby(['exit_hour', 'Direction'])['PnL'].sum().groupby('Direction').idxmax()
+    worst_hours = records.groupby(['exit_hour', 'Direction'])['PnL'].sum().groupby('Direction').idxmin()
+    
+    return {
+        'direction_stats': direction_stats,
+        'hourly_pnl': hourly_pnl,
+        'daily_pnl': daily_pnl,
+        'date_pnl': date_pnl,
+        'best_hours': best_hours,
+        'worst_hours': worst_hours
+    }
+
+# Execute analysis
+performance_analysis = analyze_trade_performance(pf)
+```
+
+## 8. Visualization Functions
+
+### Trade Performance Visualization
+```python
+def plot_trade_performance(pf):
+    """Create comprehensive trade performance plots"""
+    
+    import plotly.graph_objects as go
+    from plotly.subplots import make_subplots
+    
+    records = pf.trades.records_readable
+    records['exit_hour'] = records.index.hour
+    records['exit_day'] = records.index.day_name()
+    
+    # Create subplots
+    fig = make_subplots(
+        rows=2, cols=2,
+        subplot_titles=['Hourly P&L by Direction', 'Daily P&L by Direction', 
+                       'P&L Distribution', 'Cumulative P&L by Direction'],
+        specs=[[{"secondary_y": True}, {"secondary_y": True}],
+               [{"secondary_y": False}, {"secondary_y": False}]]
+    )
+    
+    # Hourly P&L
+    hourly_long = records[records['Direction'] == 'Long'].groupby('exit_hour')['PnL'].sum()
+    hourly_short = records[records['Direction'] == 'Short'].groupby('exit_hour')['PnL'].sum()
+    
+    fig.add_trace(go.Bar(x=hourly_long.index, y=hourly_long.values, name='Long Hourly', marker_color='green'), row=1, col=1)
+    fig.add_trace(go.Bar(x=hourly_short.index, y=hourly_short.values, name='Short Hourly', marker_color='red'), row=1, col=1)
+    
+    # Daily P&L
+    day_order = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
+    daily_long = records[records['Direction'] == 'Long'].groupby('exit_day')['PnL'].sum().reindex(day_order, fill_value=0)
+    daily_short = records[records['Direction'] == 'Short'].groupby('exit_day')['PnL'].sum().reindex(day_order, fill_value=0)
+    
+    fig.add_trace(go.Bar(x=daily_long.index, y=daily_long.values, name='Long Daily', marker_color='lightgreen'), row=1, col=2)
+    fig.add_trace(go.Bar(x=daily_short.index, y=daily_short.values, name='Short Daily', marker_color='lightcoral'), row=1, col=2)
+    
+    # P&L Distribution
+    fig.add_trace(go.Histogram(x=records[records['Direction'] == 'Long']['PnL'], name='Long Distribution', opacity=0.7), row=2, col=1)
+    fig.add_trace(go.Histogram(x=records[records['Direction'] == 'Short']['PnL'], name='Short Distribution', opacity=0.7), row=2, col=1)
+    
+    # Cumulative P&L
+    long_cumulative = records[records['Direction'] == 'Long']['PnL'].cumsum()
+    short_cumulative = records[records['Direction'] == 'Short']['PnL'].cumsum()
+    
+    fig.add_trace(go.Scatter(y=long_cumulative.values, mode='lines', name='Long Cumulative', line=dict(color='green')), row=2, col=2)
+    fig.add_trace(go.Scatter(y=short_cumulative.values, mode='lines', name='Short Cumulative', line=dict(color='red')), row=2, col=2)
+    
+    fig.update_layout(height=800, title_text="Comprehensive Trade Analysis")
+    return fig
+
+# Create visualization
+# trade_plot = plot_trade_performance(pf)
+# trade_plot.show()
+```
+
+## 9. Advanced Analytics
+
+### Trade Streaks and Patterns
+```python
+def analyze_trade_patterns(pf):
+    """Analyze trade patterns and streaks"""
+    
+    trades = pf.trades
+    records = trades.records_readable
+    
+    # Winning and losing streaks
+    winning_streaks = trades.winning_streak.records_readable
+    losing_streaks = trades.losing_streak.records_readable
+    
+    # Pattern analysis
+    patterns = {
+        'longest_winning_streak': winning_streaks['Duration'].max() if len(winning_streaks) > 0 else 0,
+        'longest_losing_streak': losing_streaks['Duration'].max() if len(losing_streaks) > 0 else 0,
+        'avg_winning_streak': winning_streaks['Duration'].mean() if len(winning_streaks) > 0 else 0,
+        'avg_losing_streak': losing_streaks['Duration'].mean() if len(losing_streaks) > 0 else 0,
+    }
+    
+    # Direction-specific patterns
+    long_patterns = analyze_direction_patterns(trades.direction_long)
+    short_patterns = analyze_direction_patterns(trades.direction_short)
+    
+    return {
+        'overall_patterns': patterns,
+        'long_patterns': long_patterns,
+        'short_patterns': short_patterns
+    }
+
+def analyze_direction_patterns(direction_trades):
+    """Analyze patterns for specific direction"""
+    
+    if direction_trades.count() == 0:
+        return {}
+    
+    return {
+        'total_trades': direction_trades.count(),
+        'win_rate': direction_trades.win_rate,
+        'profit_factor': direction_trades.profit_factor,
+        'avg_winner': direction_trades.winning.pnl.mean() if direction_trades.winning.count() > 0 else 0,
+        'avg_loser': direction_trades.losing.pnl.mean() if direction_trades.losing.count() > 0 else 0,
+        'largest_winner': direction_trades.pnl.max(),
+        'largest_loser': direction_trades.pnl.min(),
+        'total_pnl': direction_trades.pnl.sum()
+    }
+
+# Execute pattern analysis
+pattern_analysis = analyze_trade_patterns(pf)
+```
+
+## 10. Summary Report Function
+
+### Comprehensive Trade Report
+```python
+def generate_trade_report(pf):
+    """Generate comprehensive trade analysis report"""
+    
+    print("="*80)
+    print("COMPREHENSIVE TRADE ANALYSIS REPORT")
+    print("="*80)
+    
+    # Basic Statistics
+    trades = pf.trades
+    total_trades = trades.count()
+    long_trades = trades.direction_long.count()
+    short_trades = trades.direction_short.count()
+    
+    print(f"\n📊 BASIC STATISTICS")
+    print(f"Total Trades: {total_trades}")
+    print(f"Long Trades: {long_trades} ({long_trades/total_trades*100:.1f}%)")
+    print(f"Short Trades: {short_trades} ({short_trades/total_trades*100:.1f}%)")
+    
+    # P&L Analysis
+    print(f"\n💰 P&L ANALYSIS")
+    print(f"Total P&L: ${trades.pnl.sum():.2f}")
+    print(f"Long P&L: ${trades.direction_long.pnl.sum():.2f}")
+    print(f"Short P&L: ${trades.direction_short.pnl.sum():.2f}")
+    print(f"Average P&L per Trade: ${trades.pnl.mean():.2f}")
+    
+    # Temporal Analysis
+    records = trades.records_readable
+    records['exit_hour'] = records.index.hour
+    records['exit_day'] = records.index.day_name()
+    
+    print(f"\n⏰ TEMPORAL ANALYSIS")
+    
+    # Best/Worst Hours
+    hourly_pnl = records.groupby('exit_hour')['PnL'].sum()
+    best_hour = hourly_pnl.idxmax()
+    worst_hour = hourly_pnl.idxmin()
+    
+    print(f"Best Hour: {best_hour}:00 (${hourly_pnl[best_hour]:.2f})")
+    print(f"Worst Hour: {worst_hour}:00 (${hourly_pnl[worst_hour]:.2f})")
+    
+    # Best/Worst Days
+    daily_pnl = records.groupby('exit_day')['PnL'].sum()
+    best_day = daily_pnl.idxmax()
+    worst_day = daily_pnl.idxmin()
+    
+    print(f"Best Day: {best_day} (${daily_pnl[best_day]:.2f})")
+    print(f"Worst Day: {worst_day} (${daily_pnl[worst_day]:.2f})")
+    
+    # Direction Performance
+    print(f"\n📈 DIRECTION PERFORMANCE")
+    if long_trades > 0:
+        print(f"Long Win Rate: {trades.direction_long.win_rate:.2%}")
+        print(f"Long Profit Factor: {trades.direction_long.profit_factor:.2f}")
+    
+    if short_trades > 0:
+        print(f"Short Win Rate: {trades.direction_short.win_rate:.2%}")
+        print(f"Short Profit Factor: {trades.direction_short.profit_factor:.2f}")
+    
+    print("="*80)
+
+# Generate report
+generate_trade_report(pf)
+```
+
+This comprehensive analysis framework provides all the tools needed to analyze `pf.trades` with particular focus on:
+
+1. **Direction-specific analysis** - Separate analysis for long and short trades
+2. **Daily P&L patterns** - Understanding daily performance patterns
+3. **Hourly P&L by direction** - Identifying optimal trading hours for each direction
+4. **Day of week analysis** - Finding the best/worst days for different directions
+5. **Custom metrics** - Extending the analysis with domain-specific metrics
+6. **Visualization tools** - Creating comprehensive performance visualizations
+7. **Pattern recognition** - Identifying winning/losing streaks and patterns
+8. **Comprehensive reporting** - Generating detailed performance reports
+
+The framework is designed to be modular, allowing you to pick and choose the specific analyses most relevant to your trading strategy evaluation needs.