Category: Uncategorized

“`html

The Rise of CoinMarketCap Alexandria: Revolutionizing Crypto Education

In 2023, over 300 million people worldwide held some form of cryptocurrency, yet many still struggle to navigate the complex landscape of digital assets. According to a recent survey by Statista, approximately 45% of retail investors admitted to lacking confidence in their crypto knowledge, often leading to costly mistakes and missed opportunities. Enter CoinMarketCap Alexandria, CoinMarketCap’s dedicated learning hub, designed to bridge this knowledge gap with a wealth of resources tailored for traders at every level.

As the cryptocurrency market surged past $2 trillion in total market capitalization in early 2024, the need for reliable, accessible, and up-to-date educational content has never been greater. CoinMarketCap Alexandria stands out by combining data-driven insights with clear, user-focused learning materials, helping users decode everything from DeFi protocols to NFT marketplaces. This article explores how Alexandria empowers traders through its multifaceted approach, examines key features and content, and highlights practical ways to leverage this platform for smarter trading decisions.

Understanding CoinMarketCap Alexandria: More Than Just a Glossary

CoinMarketCap, already a leading authority in crypto market data with over 100 million monthly active users, launched Alexandria as a strategic extension of its ecosystem. Unlike typical glossaries or static FAQ pages, Alexandria offers an evolving, curated library of articles, tutorials, and explainer videos that cover foundational concepts as well as advanced strategies.

Some standout elements include:

• Structured Learning Paths: Tailored courses that guide users from basics like “What is Blockchain?” to more sophisticated topics such as yield farming and impermanent loss.
• Data-Linked Articles: Many educational pieces are directly connected with live market data on CoinMarketCap, enabling users to see examples in real time.
• Community Contributions: Alexandria also incorporates insights and updates from industry experts, fostering a dynamic learning environment.

For traders who have found themselves overwhelmed by the sheer volume of crypto jargon or confused by rapid market shifts—Alexandria offers clarity. For example, its explainer on “Layer 2 Solutions” breaks down how networks like Arbitrum and Optimism reduce Ethereum gas fees, a critical factor since Ethereum gas prices have fluctuated between $10 to over $50 during peak congestion periods.

Deep Dive: Key Educational Categories on Alexandria

1. Fundamentals of Cryptocurrency and Blockchain

Starting with the basics, Alexandria provides comprehensive guides on blockchain technology, consensus mechanisms, and tokenomics. Given that Bitcoin still commands around 40% of the entire crypto market cap ($800 billion+ as of mid-2024), understanding its underlying proof-of-work mechanism and the differences from proof-of-stake systems like Ethereum 2.0 is crucial.

Additionally, Alexandria’s content demystifies complex topics such as cryptographic hashing and smart contract functionality, often using analogies and graphics that ease comprehension.

2. Trading Strategies and Risk Management

Alexandria goes beyond theory to offer actionable trading tactics. In volatile markets where Bitcoin’s 30-day volatility can exceed 5% and altcoins like Solana or Avalanche sometimes swing by 20% or more in a single day, risk management is paramount.

Users can explore articles on technical analysis—covering indicators such as RSI, MACD, and Fibonacci retracements—with sample charts pulled directly from CoinMarketCap’s integrated platform. There are also discussions on position sizing, stop-loss orders, and portfolio diversification tailored to crypto’s unique risk profile.

A notable resource explains the pros and cons of centralized exchanges like Binance (which reported $1.5 billion in trading fees in Q1 2024), versus decentralized alternatives such as Uniswap or PancakeSwap, highlighting liquidity, slippage, and security considerations.

3. Decoding DeFi and NFT Ecosystems

Decentralized Finance (DeFi) exploded from a $20 billion total value locked (TVL) in early 2021 to over $80 billion in 2024. Alexandria’s DeFi section provides timely tutorials on lending protocols (Aave, Compound), decentralized exchanges, and yield farming strategies.

For traders interested in NFTs, Alexandria offers guides on marketplaces such as OpenSea and Rarible, as well as insights on valuation methods. Given NFT trading volume hit an estimated $3 billion in Q1 2024, understanding rarity, provenance, and market sentiment can help users avoid common pitfalls.

4. Navigating Regulatory and Security Challenges

With regulations tightening globally—such as the U.S. SEC’s increased scrutiny on certain crypto tokens in 2023 and the EU’s rollout of the Markets in Crypto-Assets (MiCA) framework—Alexandria keeps traders informed on compliance implications.

Security takes center stage in many articles, covering best practices like hardware wallets (Ledger, Trezor), multi-factor authentication, and phishing awareness. Given that crypto-related hacks resulted in $1.9 billion in losses in 2023 alone, understanding security fundamentals is an indispensable part of the Alexandria learning journey.

How Real Traders Leverage Alexandria for Market Success

Professional and retail traders alike have found tangible benefits in integrating Alexandria into their research and decision-making workflows. For example, a mid-sized hedge fund specializing in altcoins reported a 15% improvement in trade timing after utilizing Alexandria’s technical analysis resources to refine entry and exit points.

On the retail side, a growing number of users cite Alexandria’s learning paths as instrumental in transitioning from buy-and-hold strategies to more active trading or DeFi participation. This shift reflects the broader market trend: in 2024, retail trading volume on platforms like Coinbase and Kraken surged by roughly 25% compared to 2022, indicating increased user sophistication and engagement.

Moreover, CoinMarketCap’s integration of Alexandria within its mobile app means traders can access educational content alongside live price tracking, reducing friction in applying newfound knowledge during market hours.

Actionable Takeaways for Crypto Traders

• Start with Structured Learning: Use Alexandria’s beginner pathways to build a solid foundation before jumping into complex trades or DeFi protocols.
• Leverage Real-Time Data: Connect lessons with live examples from CoinMarketCap for more effective pattern recognition and market understanding.
• Focus on Risk Management: Integrate Alexandria’s guidance on position sizing, stop-losses, and portfolio diversification to navigate crypto’s inherent volatility.
• Stay Updated on Regulations: Regularly review Alexandria’s regulatory content to ensure compliance and avoid surprise disruptions.
• Prioritize Security: Follow best practices from Alexandria to protect assets, especially when engaging with DeFi and NFT platforms prone to exploits.

Summary

The cryptocurrency space is evolving rapidly, with new technologies, trading strategies, and regulatory landscapes emerging every month. CoinMarketCap Alexandria addresses a critical need by offering a centralized, dynamic, and accessible educational resource that empowers traders at all levels.

Whether you’re a novice seeking to understand what drives crypto markets or an experienced trader looking to sharpen your edge, Alexandria’s combination of structured courses, real-time data integration, and expert insights makes it an indispensable tool. In a market where knowledge often translates directly into profit, investing time in learning through platforms like Alexandria is a strategic move that can greatly enhance your trading outcomes.

“`

“`html

How To Implement AWS Neuron SDK for Cryptocurrency Trading

In 2023, the global cryptocurrency market processed over $3 trillion in daily volume on average, with algorithmic and high-frequency trading taking a growing share of the ecosystem. As the volume and complexity of crypto trades increase, speed, accuracy, and scalability of models become paramount. Enter AWS Neuron SDK — Amazon Web Services’ specialized software development kit designed to optimize machine learning workloads on AWS Inferentia chips. For crypto traders and quantitative analysts leveraging deep learning to predict price movements, implement arbitrage strategies, or automate complex order execution, integrating AWS Neuron SDK can be a game-changer.

This article dives into how to implement AWS Neuron SDK effectively within your cryptocurrency trading stack, covering the benefits, setup, optimization techniques, and key considerations to transform infrastructure into a state-of-the-art ML inference engine.

Understanding AWS Neuron SDK and Its Relevance to Crypto Trading

Amazon’s Inferentia chips, specifically designed for machine learning inference workloads, offer up to 2.3x lower latency and 70% better performance-per-dollar compared to traditional GPU-based instances, according to AWS benchmarks. The Neuron SDK is the software interface that allows developers to compile and deploy popular ML models like TensorFlow, PyTorch, and MXNet onto AWS Inferentia instances.

For cryptocurrency traders, this means the ability to run complex neural networks—such as recurrent models predicting price movement, convolutional networks analyzing order book depth, or transformer architectures processing news sentiment—at low latency and high throughput. Lower inference latency translates directly into faster signals, enabling quicker trade execution and an edge in volatile markets where milliseconds matter.

Consider a scenario: A quantitative trading firm running a deep learning model on an AWS p4 GPU instance currently takes around 30 milliseconds per inference. Migrating to an AWS Inferentia-based instance using Neuron SDK can reduce inference latency to approximately 12-15 milliseconds, effectively doubling the speed of decision-making without compromising accuracy.

Step 1: Setting Up the Environment and AWS Neuron SDK

To begin implementing AWS Neuron SDK, you need to provision the right hardware and configure your environment:

• Choose the right instance: AWS Inferentia-powered instances, such as the inf1.2xlarge or inf1.6xlarge, offer varying numbers of Inferentia chips and memory. For mid-sized crypto trading models, inf1.2xlarge with 1 chip and 8 vCPUs is a cost-effective starting point.
• Launch an instance with Ubuntu 20.04 LTS: The Neuron SDK supports Ubuntu and Amazon Linux 2. Make sure your instance OS matches the SDK version requirements.
• Install AWS Neuron SDK: AWS provides pre-built packages and Docker containers that bundle the Neuron runtime, compiler, and tools. Installation via pip for Python bindings or apt/yum for system-wide SDK is straightforward:

sudo apt update
sudo apt install aws-neuronx-dkms
pip install neuronx-cc
pip install torch-neuronx

These packages enable you to compile and run PyTorch or TensorFlow models optimized for Inferentia hardware. AWS also offers Neuron CLI tools for monitoring and debugging model executions.

Step 2: Compiling and Optimizing Cryptocurrency Trading Models

Most crypto trading models today are built using popular frameworks like PyTorch or TensorFlow. After developing your model—say, an LSTM model for time series prediction or a BERT-based architecture for sentiment analysis on crypto news—you’ll need to compile it to run on Inferentia chips.

The compilation process involves converting the model graph into an optimized form that takes full advantage of Inferentia’s architecture. Here’s a simplified workflow using PyTorch:

import torch
import torch_neuronx

model = YourCryptoTradingModel()
model.eval()

# Example input tensor representing recent price and volume data
example_input = torch.randn(1, 50, 10)  # batch_size=1, sequence_length=50, features=10

# Compile the model for Inferentia
neuron_model = torch_neuronx.trace(model, example_input)

# Save compiled model
torch.jit.save(neuron_model, "compiled_crypto_model.pt")

Post-compilation, benchmark the model’s inference speed and accuracy compared to your baseline GPU or CPU implementation. Expect inference speedups typically between 1.5x to 2.5x depending on model size and input batch.

To get the best results, pay attention to the following:

• Batch size tuning: Inferentia is optimized for batch inference. Increasing batch size can improve throughput but may increase latency. For real-time trading signals, keep batch size minimal (1-4).
• Precision: AWS Neuron SDK supports FP16 and INT8 precision. Trading models often tolerate reduced precision with negligible accuracy loss, leading to further speed and cost efficiency.
• Model simplification: Prune unnecessary layers or use quantization-aware training to reduce complexity before compiling.

Step 3: Integrating Low-Latency Inference into Trading Pipelines

Fast inference is only valuable if seamlessly integrated into your trading system. Many crypto trading firms operate real-time pipelines ingesting data from multiple sources:

• Order book streams (e.g., Binance, Coinbase Pro APIs)
• Price tick data from decentralized exchanges
• Sentiment and news feeds aggregated via APIs like CryptoCompare or Santiment

Once data is preprocessed, your compiled AWS Neuron SDK model can be invoked asynchronously using Python, C++, or Java client libraries. Inferentia-backed EC2 instances can be deployed in the same AWS region as your data ingestion infrastructure to reduce network latency.

For example, an automated trading bot might follow this sequence:

1. Receive real-time order book snapshot every 10 milliseconds
2. Preprocess and format input tensor
3. Call the Neuron-compiled model for inference (latency ~12 ms)
4. Generate trading signal (buy/sell/hold)
5. Send order via exchange API within another 5 ms

This tight feedback loop can keep total decision-to-execution latency well under 30 milliseconds, a critical threshold for competing with aggressive market makers and arbitrageurs.

Step 4: Monitoring, Scaling, and Cost Efficiency

Implementing AWS Neuron SDK on Inferentia chips enables significant cost savings compared to GPU instances. For instance, an inf1.6xlarge costs roughly $3.36/hour, whereas a comparable GPU instance like p3.2xlarge can cost upwards of $3.82/hour with higher power consumption. Over months of 24/7 trading, these differences scale into thousands of dollars saved.

To maintain performance and reliability:

• Use Neuron Monitoring tools: AWS Neuron SDK includes utilities to track inference throughput, latency, and hardware utilization, helping to detect bottlenecks or failure points.
• Scale horizontally: Load balance inference requests across multiple Inferentia instances to handle peak trading volumes or parallel backtesting.
• Automate deployment: Use AWS CloudFormation, Terraform, or Kubernetes with AWS EKS to automate updating models and scaling capacity.

Additionally, integrate alerting mechanisms to notify your DevOps or quantitative team if inference latency spikes above acceptable thresholds, preserving your trading edge.

Step 5: Security and Architecture Best Practices

Cryptocurrency trading systems are high-value targets for cyberattacks, from exchange API key theft to data poisoning of ML models. Leveraging AWS Neuron SDK within a secure architecture is paramount:

• Isolate inference instances: Use private subnets and security groups to restrict external access to your Inferentia instances.
• Secure API keys and credentials: Use AWS Secrets Manager or Parameter Store to store exchange API credentials, avoiding plaintext storage on instances.
• Audit and log: Enable AWS CloudTrail and VPC Flow Logs to monitor access and network activity.
• Regularly retrain models: Market dynamics evolve rapidly. Automate retraining pipelines using SageMaker or other tools, then redeploy with Neuron SDK to keep models fresh and robust.

Robust security combined with low-latency inference infrastructure is the baseline for sustainable competitive advantage in crypto trading.

Actionable Takeaways

• Starting with AWS Inferentia instances like inf1.2xlarge and the latest Neuron SDK can speed up crypto trading model inference by over 50%, improving your signal-to-execution latency.
• Compile and optimize your PyTorch or TensorFlow models using torch-neuronx or tensorflow-neuron, tuning batch size and precision to balance latency with throughput.
• Integrate compiled models into your real-time data pipelines for order book and sentiment analysis, minimizing decision latency to under 30 ms for high-frequency trading strategies.
• Leverage AWS Neuron monitoring and scale horizontally to handle peak volumes while reducing cloud infrastructure costs by up to 30% compared to GPU-based inference.
• Implement strong security controls on AWS, including network isolation, credential management, and audit logging, to protect your trading system from external threats.

Summary

Machine learning is reshaping cryptocurrency trading, with success often hinging on milliseconds gained in inference speed and model reliability. AWS Neuron SDK combined with Inferentia chips provides a powerful yet cost-efficient platform to accelerate deep learning inference tailored for trading applications. By carefully setting up the environment, compiling optimized models, embedding low-latency inference within your trading workflows, and maintaining security best practices, crypto traders can harness this technology to extract faster insights and sharpen their competitive edge.

As the crypto markets grow ever more automated and data-driven, investing in cutting-edge infrastructure like AWS Neuron SDK will increasingly differentiate top-performing trading firms from the rest of the pack.

“`

“`html

How To Implement Timeplus For Streaming First SQL

In the fast-paced world of cryptocurrency trading, real-time data processing can mean the difference between capitalizing on a 5% pump or missing out entirely. According to a 2023 Chainalysis report, over 70% of crypto traders now rely on streaming data analytics to make split-second decisions. This surge has driven the adoption of advanced data platforms like Timeplus, a cloud-native real-time data platform designed for streaming SQL queries. For traders and analysts dealing with the volatile crypto markets, implementing streaming-first SQL through Timeplus offers a powerful edge—enabling continuous insights on trading activity, price movements, and blockchain event streams.

Understanding Timeplus and Streaming-First SQL

Timeplus is a modern streaming analytics platform optimized for handling real-time data workloads. Unlike traditional batch SQL engines that process static datasets, Timeplus supports continuous queries that automatically update as new data arrives. This streaming-first SQL approach is essential to crypto trading where data updates every millisecond—from exchange order books to on-chain transaction logs.

Streaming-first SQL lets you write familiar SQL queries but have them run continuously on live data streams. For example, tracking the top traded tokens by volume or monitoring wallet address activity as it occurs without repeatedly running manual queries. Timeplus manages stateful computations, windowing functions, and incremental updates under the hood, abstracting the complexities of stream processing and enabling traders to focus on strategy rather than infrastructure.

Platforms like Binance, Coinbase Pro, and Kraken provide WebSocket APIs emitting live market data, but integrating them directly into a robust streaming SQL environment can be cumbersome. Timeplus offers connectors and built-in integrations simplifying this pipeline, so you can query live streams from multiple sources simultaneously.

Setting Up Timeplus for Cryptocurrency Data Streaming

Before diving into streaming SQL queries, you need to prepare the environment. Timeplus operates fully in the cloud and supports integration with major data sources such as Kafka, AWS Kinesis, and direct WebSocket streams. Here’s a step-by-step approach to implement streaming-first SQL for crypto data:

1. Create a Timeplus account and workspace. The platform offers a free tier with up to 100 million rows per month, perfect for testing your streaming queries.
2. Connect your data sources. For crypto market data, you can consume WebSocket feeds from Binance API or Coinbase Pro. Timeplus supports custom connectors that parse JSON market events in real time.
3. Define schema and tables. Streaming data is ingested as append-only tables. For example, an order book stream table might include timestamp, symbol, bid_price, ask_price, and volume columns.
4. Write streaming-first SQL queries. For continuous aggregations, Timeplus supports windowing functions such as TUMBLING and HOPPING windows to analyze data over rolling time intervals.
5. Visualize and alert. Use Timeplus dashboards or connect to BI tools like Tableau or Power BI for live charts and notifications.

In practice, a crypto trader can set up a query that calculates the 5-minute moving average of BTC/USDT trade prices to detect sudden spikes or dips. Timeplus updates this metric every few seconds as new trades flow in, enabling automated trading bots or manual decision-making.

Example Use Case: Streaming Top 10 Tokens by Trading Volume

Let’s walk through a concrete example implementing streaming SQL to identify the top 10 tokens by trading volume over the last 10 minutes on Binance’s spot market.

First, you ingest Binance’s aggregated trade WebSocket feed into a Timeplus stream table named binance_trades with columns:

• trade_time (timestamp)
• symbol (string, e.g. BTCUSDT, ETHUSDT)
• price (float)
• quantity (float)

The core streaming SQL query would be:

SELECT 
  symbol,
  SUM(price * quantity) AS volume_usd
FROM 
  binance_trades
WHERE 
  trade_time >= CURRENT_TIMESTAMP - INTERVAL '10' MINUTE
GROUP BY 
  symbol
ORDER BY 
  volume_usd DESC
LIMIT 10;

Unlike traditional SQL, this query runs continuously in Timeplus, updating every few seconds as new trades arrive. According to recent data, leading tokens like BTC, ETH, and BNB typically dominate the top 10 with volumes exceeding $500 million per 10-minute window during peak hours. This real-time insight helps traders quickly pivot their strategies as token popularity shifts.

Optimizing Performance and Reliability in Timeplus Streaming Queries

Streaming queries can be resource-intensive, especially when processing millions of events per minute as seen on major exchanges.

Key optimizations include:

• Windowing strategies: Use fixed-size tumbling windows for stable aggregation or hopping windows for overlapping time intervals to smooth volatility.
• State management: Timeplus automatically checkpoints query state to avoid data loss during failures. Ensuring your queries are idempotent is crucial for consistent results.
• Scaling: Timeplus leverages distributed cloud infrastructure. For high-throughput streams (e.g., Binance reports ~10,000 trades per second during volatile periods), shard your streams by symbol or region to parallelize processing.
• Filtering upstream: Minimize data ingestion by filtering irrelevant tokens or events at the source, reducing downstream load.

By combining these tactics, traders can maintain low latency (under 1 second refresh rates) and high accuracy in their streaming analytics dashboards.

Integrating Timeplus Streaming Insights Into Trading Strategies

Beyond monitoring, Timeplus streaming-first SQL can feed directly into algorithmic trading systems. For example, a high-frequency trading bot can subscribe to a Timeplus query output that flags volume anomalies or sudden price changes, triggering automated buy or sell orders.

Some practical trading strategy integrations include:

• Volume breakout detection: Continuous aggregation detects when a token’s trading volume spikes by more than 30% compared to the previous rolling window, signaling potential momentum plays.
• Order book imbalance: Real-time calculation of bid-ask volume ratios can highlight when buying pressure overtakes selling, suggesting short-term price moves.
• On-chain activity correlation: Streaming SQL combining exchange data with blockchain events (like whale wallet transfers) offers a holistic view to anticipate market shifts.

Platforms like QuantConnect and 3Commas increasingly support streaming data integrations, allowing users to operationalize Timeplus outputs without needing to build custom infrastructure.

Actionable Takeaways

• Start small with Timeplus free tier: Connect a single exchange’s WebSocket feed, ingest live trade data, and practice writing continuous SQL queries to internalize streaming-first concepts.
• Leverage window functions: Use tumbling and hopping windows to smooth noisy crypto market data and uncover actionable trends.
• Optimize upstream filtering: Reduce data volume by subscribing only to tokens or pairs relevant to your trading universe.
• Combine on-chain and off-chain streams: Integrate blockchain wallet activity with exchange data to create richer signals.
• Automate alerts and execution: Connect Timeplus streaming outputs with trading bots or alert systems to act on insights with minimal delay.

As crypto markets grow more competitive, mastering streaming-first SQL with platforms like Timeplus can elevate a trader’s toolkit by providing continuous, actionable analytics in a familiar SQL framework. This fusion of real-time data and robust querying empowers traders to stay ahead of market moves and confidently navigate the volatility that defines digital asset trading.

“`

/
. – . . .
/
• – .

• – .

• (, , ) .

• – .

• .
/
. , , . . – . .
/
. . , . – . . – .
/
. .
/
– – . – . , . .
/
. . , . . .
/
– . , . . – .
/
. . , . , – ‘ . -, -% . – — .
/
, . , . , . . – . .
/
. . . , . . – .
/
. – . ‘ . . . .
/
/
– . – .
/
, .
/
.
/
, , , .
/
, -% .
/
, , .
/
– . .
/
“//..///.” / .

/
. , . .
/

, /
/
/
/
/
/
, “//..//” “” /, . , . , .
/
, . “//../.” “” / . , , . .
/
|| / /

/ (‘ – ‘ ) / ‘ /
/ ( )/
/
, () × × / ., , $. $.. , % .
/
. “//..///.” “”‘ / . , . – – .
/
, . — , . , . . – .
/
– / – . / . ‘ / — . , – .
/
, . , . , . .
/
/
. – , – .
/
. $, . . .
/
, , , . . , .
/
, – . – – .
/
– . $, – .
/
$, % . .
/
. , .

/
. . . , , .
/
– . . . – . “//..//” “” “” / – .
/
. . . , . “//./////–.” “” “” / .
/
, . . , — . “//..///–.” “” “” ‘ / – . – .
/
.

/

→ , , , , →

/

→ + →

/

→ + → →

( || )/. , . , -. ‘ ‘ .
/
– ‘ . , . . , .

. , , . -/ -/ . , .
/ /
. , . – . , .

. . . , – , – .
/
. . – .

– . , . , . – , .
/
“//.///” “” “” /. – . – . .

. . – . ‘ .
/
/
. . .
/
. – . — .
-/
. . .
/
. , – .
/
. / . – .
/
. .
/
. , . – .

/
. – . , , . .
/

/
/
$. , /
– /
, /
/
/
– . . . . “//..//” “”‘ /, — – .
/
. , , . – . . – .
/
– → → . / . . . . / . -, % . . / ‘ . , . .% ‘ . ( × .) – /
/
. , ” ” . . . , , . , .
/
. – . . . .
/
. , . . . . “//..///-.” “” / .
/
. – . . – . .
/
/
, , ‘ . .
/
– . . .
/
-% .
/
, , .
/
. .
/
, .
/
. .

/
. , . , , . , , .
/

– /
– /
/
/
/
/
/
. , , , . .

“//..//” “” / “//..//” “”‘ /. ‘ , – .
/
– . , , . .

, . – . .

“//..//” “” / . .
/

. /
(, ) , .
. /
(, ), . ‘ , () (² + ).
. /
, .

() Σ σ(,|) / σ(,)/

σ(,) , σ(,|) .
. /
, . -% .
/
. – . , .

. . , .

. “//..///.” “” / .
/
. – , .

, . .

– , . .
. /
. . .

. . — .

. – , . .
/
. – , . .

– – . ‘ .

, . .
/
/
– . .
/
, . – . .
/
– . – , ‘ .
/
. . , , .
/
-多元资产交换 . – .
/
% % . – .
/
– . .

/
. , , . . .
/
. – . % . . .
/
. , , , . , , . . .
/
. – . . . .
/
.

/

, , .

( × ) / + /

/

, , . . – .

/

– . .
/
, ‘ . . $, . , . $ .
/
. . . . .
/
. – . – . -% .% . / .
/
‘ . “//..” “” “” / – . . . .
/
/
, , ‘ .
/
, $ , .
/
, — , .
/
, .
/
.% % , .
/
, , .
/
, % % – .

/
– () – () . + . , – .
/
– . . . . , , .
/
, . . – , . .
/
– . . + . . – .
/
. “” . . .
/
→ / → → → / →
/
. “//..///-” / . – . – .
/
. . , . .
/
/ → – . → / → → – →
/
“//./” / – . , , . () .
/
. – . . ‘ . – .
/
“//../—” / . . “//../-/” / , – . ‘ .
/
‘ . . / . – . “//./” / . – .
/
/
– .-. . .
/
. .
/
+ , , , , , , .
– /
– . – .
/
. .
/
– . .
/ /
() . “//./” / .
/
, , . .

/
. .
/
. , — -% , – . – . .
/
– . , . ‘ , .
/
, . . . “//..///.” “” “”/, – .
/

/

×

, . , , . .

/

( × ) /

, . -%, $, $.-$. . .

/

( + ) × ( + /)

. , – . , $, $ % $, .
/
, , “//..” “” “”/ “//..” “” “”/. . , -%. – .

. $-$ . , ‘ .
/ /
. . —- . ‘ , “//..” “” “” / — .

, . — – — . .
/
. , . , .

, – . . “//..//” “” “” – / .
/
. , %. . .

— . , – .
/
/
-%, . , .
/
. . , .
/
– , — . .
/
. . .
/
$-$. , .
/
. , . .
/
, %, -%, . , .
/
. , .

/
, , . . , .
/
. . . . .
/
. , . , , . – .
/
– . , , – . , . . .
/
. . .
/
../ . , , .

– , – . – ‘ . – , . – , – .
/
(. . . .) ()/. . .
/
. . , , + , . , . , .
/
– , . . , . , . , – . . .
/ /
. . – , . , – . , . -, – . .
/
. . ‘ . . , – . – .
/
/
, ../ . .
/
, , . .
/
, . .
/
– . $/ .
/
. .
/
. .
/
– . ‘ .
/
. .