Python Volatility-Based Catapult Indicator

As a catapult, this technical indicator uses three systems: Volatility (the fulcrum), Momentum (the propeller), and a Directional Filter (Acting as the support). The goal is to get a signal that predicts volatility acceleration and direction based on historical patterns. We want to know when the market will move. and where. This indicator outperforms standard indicators.

Knowledge must be accessible to everyone. This is why my new publications Contrarian Trading Strategies in Python and Trend Following Strategies in Python now include free PDF copies of my first three books (Therefore, purchasing one of the new books gets you 4 books in total). GitHub-hosted advanced indications and techniques are in the two new books above.

The Foundation: Volatility

The Catapult predicts significant changes with the 21-period Relative Volatility Index.

The Average True Range, Mean Absolute Deviation, and Standard Deviation all assess volatility. Standard Deviation will construct the Relative Volatility Index.

Standard Deviation is the most basic volatility. It underpins descriptive statistics and technical indicators like Bollinger Bands. Before calculating Standard Deviation, let's define Variance.

Variance is the squared deviations from the mean (a dispersion measure). We take the square deviations to compel the distance from the mean to be non-negative, then we take the square root to make the measure have the same units as the mean, comparing apples to apples (mean to standard deviation standard deviation). Variance formula:

As stated, standard deviation is:

# The function to add a number of columns inside an array
def adder(Data, times):
    
    for i in range(1, times + 1):
    
        new_col = np.zeros((len(Data), 1), dtype = float)
        Data = np.append(Data, new_col, axis = 1)
        
    return Data

# The function to delete a number of columns starting from an index
def deleter(Data, index, times):
    
    for i in range(1, times + 1):
    
        Data = np.delete(Data, index, axis = 1)
        
    return Data
    
# The function to delete a number of rows from the beginning
def jump(Data, jump):
    
    Data = Data[jump:, ]
    
    return Data

# Example of adding 3 empty columns to an array
my_ohlc_array = adder(my_ohlc_array, 3)

# Example of deleting the 2 columns after the column indexed at 3
my_ohlc_array = deleter(my_ohlc_array, 3, 2)

# Example of deleting the first 20 rows
my_ohlc_array = jump(my_ohlc_array, 20)

# Remember, OHLC is an abbreviation of Open, High, Low, and Close and it refers to the standard historical data file

def volatility(Data, lookback, what, where):
    
  for i in range(len(Data)):

     try:

        Data[i, where] = (Data[i - lookback + 1:i + 1, what].std())
     except IndexError:
        pass
        
  return Data

The RSI is the most popular momentum indicator, and for good reason—it excels in range markets. Its 0–100 range simplifies interpretation. Fame boosts its potential.

The more traders and portfolio managers look at the RSI, the more people will react to its signals, pushing market prices. Technical Analysis is self-fulfilling, therefore this theory is obvious yet unproven.

RSI is determined simply. Start with one-period pricing discrepancies. We must remove each closing price from the previous one. We then divide the smoothed average of positive differences by the smoothed average of negative differences. The RSI algorithm converts the Relative Strength from the last calculation into a value between 0 and 100.

def ma(Data, lookback, close, where): 
    
    Data = adder(Data, 1)
    
    for i in range(len(Data)):
           
            try:
                Data[i, where] = (Data[i - lookback + 1:i + 1, close].mean())
            
            except IndexError:
                pass
            
    # Cleaning
    Data = jump(Data, lookback)
    
    return Data
def ema(Data, alpha, lookback, what, where):
    
    alpha = alpha / (lookback + 1.0)
    beta  = 1 - alpha
    
    # First value is a simple SMA
    Data = ma(Data, lookback, what, where)
    
    # Calculating first EMA
    Data[lookback + 1, where] = (Data[lookback + 1, what] * alpha) + (Data[lookback, where] * beta)    
 
    # Calculating the rest of EMA
    for i in range(lookback + 2, len(Data)):
            try:
                Data[i, where] = (Data[i, what] * alpha) + (Data[i - 1, where] * beta)
        
            except IndexError:
                pass
            
    return Datadef rsi(Data, lookback, close, where, width = 1, genre = 'Smoothed'):
    
    # Adding a few columns
    Data = adder(Data, 7)
    
    # Calculating Differences
    for i in range(len(Data)):
        
        Data[i, where] = Data[i, close] - Data[i - width, close]
     
    # Calculating the Up and Down absolute values
    for i in range(len(Data)):
        
        if Data[i, where] > 0:
            
            Data[i, where + 1] = Data[i, where]
            
        elif Data[i, where] < 0:
            
            Data[i, where + 2] = abs(Data[i, where])
            
    # Calculating the Smoothed Moving Average on Up and Down
    absolute values        
                             
    lookback = (lookback * 2) - 1 # From exponential to smoothed
    Data = ema(Data, 2, lookback, where + 1, where + 3)
    Data = ema(Data, 2, lookback, where + 2, where + 4)
    
    # Calculating the Relative Strength
    Data[:, where + 5] = Data[:, where + 3] / Data[:, where + 4]
    
    # Calculate the Relative Strength Index
    Data[:, where + 6] = (100 - (100 / (1 + Data[:, where + 5])))  
    
    # Cleaning
    Data = deleter(Data, where, 6)
    Data = jump(Data, lookback)

    return Data

def relative_volatility_index(Data, lookback, close, where):

    # Calculating Volatility
    Data = volatility(Data, lookback, close, where)
    
    # Calculating the RSI on Volatility
    Data = rsi(Data, lookback, where, where + 1) 
    
    # Cleaning
    Data = deleter(Data, where, 1)
    
    return Data

The Arm Section: Speed

The Catapult predicts momentum direction using the 14-period Relative Strength Index.

As a reminder, the RSI ranges from 0 to 100. Two levels give contrarian signals:

• A positive response is anticipated when the market is deemed to have gone too far down at the oversold level 30, which is 30.

• When the market is deemed to have gone up too much, at overbought level 70, a bearish reaction is to be expected.

Comparing the RSI to 50 is another intriguing use. RSI above 50 indicates bullish momentum, while below 50 indicates negative momentum.

The direction-finding filter in the frame

The Catapult's directional filter uses the 200-period simple moving average to keep us trending. This keeps us sane and increases our odds.

Moving averages confirm and ride trends. Its simplicity and track record of delivering value to analysis make them the most popular technical indicator. They help us locate support and resistance, stops and targets, and the trend. Its versatility makes them essential trading tools.

This is the plain mean, employed in statistics and everywhere else in life. Simply divide the number of observations by their total values. Mathematically, it's:

We defined the moving average function above. Create the Catapult indication now.

Indicator of the Catapult

The indicator is a healthy mix of the three indicators:

• The first trigger will be provided by the 21-period Relative Volatility Index, which indicates that there will now be above average volatility and, as a result, it is possible for a directional shift.

• If the reading is above 50, the move is likely bullish, and if it is below 50, the move is likely bearish, according to the 14-period Relative Strength Index, which indicates the likelihood of the direction of the move.

• The likelihood of the move's direction will be strengthened by the 200-period simple moving average. When the market is above the 200-period moving average, we can infer that bullish pressure is there and that the upward trend will likely continue. Similar to this, if the market falls below the 200-period moving average, we recognize that there is negative pressure and that the downside is quite likely to continue.

lookback_rvi = 21
lookback_rsi = 14
lookback_ma  = 200
my_data = ma(my_data, lookback_ma, 3, 4)
my_data = rsi(my_data, lookback_rsi, 3, 5)
my_data = relative_volatility_index(my_data, lookback_rvi, 3, 6)

Two-handled overlay indicator Catapult. The first exhibits blue and green arrows for a buy signal, and the second shows blue and red for a sell signal.

The chart below shows recent EURUSD hourly values.

def signal(Data, rvi_col, signal):
    
    Data = adder(Data, 10)
        
    for i in range(len(Data)):
            
        if Data[i,     rvi_col] < 30 and \
           Data[i - 1, rvi_col] > 30 and \
           Data[i - 2, rvi_col] > 30 and \
           Data[i - 3, rvi_col] > 30 and \
           Data[i - 4, rvi_col] > 30 and \
           Data[i - 5, rvi_col] > 30:
               
               Data[i, signal] = 1
                           
    return Data

Signals are straightforward. The indicator can be utilized with other methods.

my_data = signal(my_data, 6, 7)

Lumiwealth shows how to develop all kinds of algorithms. I recommend their hands-on courses in algorithmic trading, blockchain, and machine learning.

Summary

To conclude, my goal is to contribute to objective technical analysis, which promotes more transparent methods and strategies that must be back-tested before implementation. Technical analysis will lose its reputation as subjective and unscientific.

After you find a trading method or approach, follow these steps:

• Put emotions aside and adopt an analytical perspective.

• Test it in the past in conditions and simulations taken from real life.

• Try improving it and performing a forward test if you notice any possibility.

• Transaction charges and any slippage simulation should always be included in your tests.

• Risk management and position sizing should always be included in your tests.

After checking the aforementioned, monitor the plan because market dynamics may change and render it unprofitable.

Python Heatmap Technical Indicator

Heatmaps provide an instant overview. They can be used with correlations or to predict reactions or confirm the trend in trading. This article covers RSI heatmap creation.

The Market System

Market regime:

• Bullish trend: The market tends to make higher highs, which indicates that the overall trend is upward.

• Sideways: The market tends to fluctuate while staying within predetermined zones.

• Bearish trend: The market has the propensity to make lower lows, indicating that the overall trend is downward.

Most tools detect the trend, but we cannot predict the next state. The best way to solve this problem is to assume the current state will continue and trade any reactions, preferably in the trend.

If the EURUSD is above its moving average and making higher highs, a trend-following strategy would be to wait for dips before buying and assuming the bullish trend will continue.

Indicator of Relative Strength

J. Welles Wilder Jr. introduced the RSI, a popular and versatile technical indicator. Used as a contrarian indicator to exploit extreme reactions. Calculating the default RSI usually involves these steps:

• Determine the difference between the closing prices from the prior ones.

• Distinguish between the positive and negative net changes.

• Create a smoothed moving average for both the absolute values of the positive net changes and the negative net changes.

• Take the difference between the smoothed positive and negative changes. The Relative Strength RS will be the name we use to describe this calculation.

• To obtain the RSI, use the normalization formula shown below for each time step.

The 13-period RSI and black GBPUSD hourly values are shown above. RSI bounces near 25 and pauses around 75. Python requires a four-column OHLC array for RSI coding.

import numpy as np
def add_column(data, times):
    
    for i in range(1, times + 1):
    
        new = np.zeros((len(data), 1), dtype = float)
        
        data = np.append(data, new, axis = 1)
    return data
def delete_column(data, index, times):
    
    for i in range(1, times + 1):
    
        data = np.delete(data, index, axis = 1)
    return data
def delete_row(data, number):
    
    data = data[number:, ]
    
    return data
def ma(data, lookback, close, position): 
    
    data = add_column(data, 1)
    
    for i in range(len(data)):
           
            try:
                
                data[i, position] = (data[i - lookback + 1:i + 1, close].mean())
            
            except IndexError:
                
                pass
            
    data = delete_row(data, lookback)
    
    return data
def smoothed_ma(data, alpha, lookback, close, position):
    
    lookback = (2 * lookback) - 1
    
    alpha = alpha / (lookback + 1.0)
    
    beta  = 1 - alpha
    
    data = ma(data, lookback, close, position)
    data[lookback + 1, position] = (data[lookback + 1, close] * alpha) + (data[lookback, position] * beta)
    for i in range(lookback + 2, len(data)):
        
            try:
                
                data[i, position] = (data[i, close] * alpha) + (data[i - 1, position] * beta)
        
            except IndexError:
                
                pass
            
    return data
def rsi(data, lookback, close, position):
    
    data = add_column(data, 5)
    
    for i in range(len(data)):
        
        data[i, position] = data[i, close] - data[i - 1, close]
     
    for i in range(len(data)):
        
        if data[i, position] > 0:
            
            data[i, position + 1] = data[i, position]
            
        elif data[i, position] < 0:
            
            data[i, position + 2] = abs(data[i, position])
            
    data = smoothed_ma(data, 2, lookback, position + 1, position + 3)
    data = smoothed_ma(data, 2, lookback, position + 2, position + 4)
    data[:, position + 5] = data[:, position + 3] / data[:, position + 4]
    
    data[:, position + 6] = (100 - (100 / (1 + data[:, position + 5])))
    data = delete_column(data, position, 6)
    data = delete_row(data, lookback)
    return data

Make sure to focus on the concepts and not the code. You can find the codes of most of my strategies in my books. The most important thing is to comprehend the techniques and strategies.

My weekly market sentiment report uses complex and simple models to understand the current positioning and predict the future direction of several major markets. Check out the report here:

Using the Heatmap to Find the Trend

RSI trend detection is easy but useless. Bullish and bearish regimes are in effect when the RSI is above or below 50, respectively. Tracing a vertical colored line creates the conditions below. How:

• When the RSI is higher than 50, a green vertical line is drawn.

• When the RSI is lower than 50, a red vertical line is drawn.

Zooming out yields a basic heatmap, as shown below.

Plot code:

def indicator_plot(data, second_panel, window = 250):
    fig, ax = plt.subplots(2, figsize = (10, 5))
    sample = data[-window:, ]
    for i in range(len(sample)):
        ax[0].vlines(x = i, ymin = sample[i, 2], ymax = sample[i, 1], color = 'black', linewidth = 1)  
        if sample[i, 3] > sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 0], ymax = sample[i, 3], color = 'black', linewidth = 1.5)  
        if sample[i, 3] < sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 3], ymax = sample[i, 0], color = 'black', linewidth = 1.5)  
        if sample[i, 3] == sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 3], ymax = sample[i, 0], color = 'black', linewidth = 1.5)  
    ax[0].grid() 
    for i in range(len(sample)):
        if sample[i, second_panel] > 50:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'green', linewidth = 1.5)  
        if sample[i, second_panel] < 50:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'red', linewidth = 1.5)  
    ax[1].grid()
indicator_plot(my_data, 4, window = 500)

Call RSI on your OHLC array's fifth column. 4. Adjusting lookback parameters reduces lag and false signals. Other indicators and conditions are possible.

Another suggestion is to develop an RSI Heatmap for Extreme Conditions.

Contrarian indicator RSI. The following rules apply:

• Whenever the RSI is approaching the upper values, the color approaches red.

• The color tends toward green whenever the RSI is getting close to the lower values.

Zooming out yields a basic heatmap, as shown below.

Plot code:

import matplotlib.pyplot as plt
def indicator_plot(data, second_panel, window = 250):
    fig, ax = plt.subplots(2, figsize = (10, 5))
    sample = data[-window:, ]
    for i in range(len(sample)):
        ax[0].vlines(x = i, ymin = sample[i, 2], ymax = sample[i, 1], color = 'black', linewidth = 1)  
        if sample[i, 3] > sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 0], ymax = sample[i, 3], color = 'black', linewidth = 1.5)  
        if sample[i, 3] < sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 3], ymax = sample[i, 0], color = 'black', linewidth = 1.5)  
        if sample[i, 3] == sample[i, 0]:
            ax[0].vlines(x = i, ymin = sample[i, 3], ymax = sample[i, 0], color = 'black', linewidth = 1.5)  
    ax[0].grid() 
    for i in range(len(sample)):
        if sample[i, second_panel] > 90:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'red', linewidth = 1.5)  
        if sample[i, second_panel] > 80 and sample[i, second_panel] < 90:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'darkred', linewidth = 1.5)  
        if sample[i, second_panel] > 70 and sample[i, second_panel] < 80:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'maroon', linewidth = 1.5)  
        if sample[i, second_panel] > 60 and sample[i, second_panel] < 70:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'firebrick', linewidth = 1.5) 
        if sample[i, second_panel] > 50 and sample[i, second_panel] < 60:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'grey', linewidth = 1.5) 
        if sample[i, second_panel] > 40 and sample[i, second_panel] < 50:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'grey', linewidth = 1.5) 
        if sample[i, second_panel] > 30 and sample[i, second_panel] < 40:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'lightgreen', linewidth = 1.5)
        if sample[i, second_panel] > 20 and sample[i, second_panel] < 30:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'limegreen', linewidth = 1.5) 
        if sample[i, second_panel] > 10 and sample[i, second_panel] < 20:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'seagreen', linewidth = 1.5)  
        if sample[i, second_panel] > 0 and sample[i, second_panel] < 10:
            ax[1].vlines(x = i, ymin = 0, ymax = 100, color = 'green', linewidth = 1.5)
    ax[1].grid()
indicator_plot(my_data, 4, window = 500)

Dark green and red areas indicate imminent bullish and bearish reactions, respectively. RSI around 50 is grey.

Summary

To conclude, my goal is to contribute to objective technical analysis, which promotes more transparent methods and strategies that must be back-tested before implementation.

Technical analysis will lose its reputation as subjective and unscientific.

When you find a trading strategy or technique, follow these steps:

• Put emotions aside and adopt a critical mindset.

• Test it in the past under conditions and simulations taken from real life.

• Try optimizing it and performing a forward test if you find any potential.

• Transaction costs and any slippage simulation should always be included in your tests.

• Risk management and position sizing should always be considered in your tests.

After checking the above, monitor the strategy because market dynamics may change and make it unprofitable.

What quarterly bank earnings reveal

Big banks know the economy best. Unless we’re talking about a housing crisis in 2007…

Banks are crucial to the U.S. economy. The Fed, communities, and investments exchange money.

An economy depends on money flow. Banks' views on the economy can affect their decision-making.

Most large banks released quarterly earnings and forward guidance last week. Others were pessimistic about the future.

What Makes Banks Confident

Bank of America's profit decreased 30% year-over-year, but they're optimistic about the economy. Comparatively, they're bullish.

Who banks serve affects what they see. Bank of America supports customers.

They think consumers' future is bright. They believe this for many reasons.

The average customer has decent credit, unless the system is flawed. Bank of America's new credit card and mortgage borrowers averaged 771. New-car loan and home equity borrower averages were 791 and 797.

2008's housing crisis affected people with scores below 620.

Bank of America and the economy benefit from a robust consumer. Major problems can be avoided if individuals maintain spending.

Reasons Other Banks Are Less Confident

Spending requires income. Many companies, mostly in the computer industry, have announced they will slow or freeze hiring. Layoffs are frequently an indication of poor times ahead.

BOA is positive, but investment banks are bearish.

Jamie Dimon, CEO of JPMorgan, outlined various difficulties our economy could confront.

But geopolitical tension, high inflation, waning consumer confidence, the uncertainty about how high rates have to go and the never-before-seen quantitative tightening and their effects on global liquidity, combined with the war in Ukraine and its harmful effect on global energy and food prices are very likely to have negative consequences on the global economy sometime down the road.

That's more headwinds than tailwinds.

JPMorgan, which helps with mergers and IPOs, is less enthusiastic due to these concerns. Incoming headwinds signal drying liquidity, they say. Less business will be done.

Final Reflections

I don't think we're done. Yes, stocks are up 10% from a month ago. It's a long way from old highs.

I don't think the stock market is a strong economic indicator.

Many executives foresee a 2023 recession. According to the traditional definition, we may be in a recession when Q2 GDP statistics are released next week.

Regardless of criteria, I predict the economy will have a terrible year.

Weekly layoffs are announced. Inflation persists. Will prices return to 2020 levels if inflation cools? Perhaps. Still expensive energy. Ukraine's war has global repercussions.

I predict BOA's next quarter earnings won't be as bullish about the consumer's strength.

Let's see this once-dominant IT corporation crumble.

What's the capital of Kazakhstan? If you don't know the answer, you can probably find it by Googling. Google Search returned results for Nur-Sultan in 0.66 seconds.

Google is the best search engine I've ever used. Did you know another search engine ruled the Internet? I'm sure you guessed Yahoo!

Google's friendly UI and wide selection of services make it my top choice. Let's explore Yahoo's decline.

Yahoo!

YAHOO stands for Yet Another Hierarchically Organized Oracle. Jerry Yang and David Filo established Yahoo.

Yahoo is primarily a search engine and email provider. It offers News and an advertising platform. It was a popular website in 1995 that let people search the Internet directly. Yahoo began offering free email in 1997 by acquiring RocketMail.

According to a study, Yahoo used Google Search Engine technology until 2000 and then developed its own in 2004.

Yahoo! rejected buying Google for $1 billion

Larry Page and Sergey Brin, Google's founders, approached Yahoo in 1998 to sell Google for $1 billion so they could focus on their studies. Yahoo denied the offer, thinking it was overvalued at the time.

Yahoo realized its error and offered Google $3 billion in 2002, but Google demanded $5 billion since it was more valuable. Yahoo thought $5 billion was overpriced for the existing market.

In 2022, Google is worth $1.56 Trillion.

What happened to Yahoo!

Yahoo refused to buy Google, and Google's valuation rose, making a purchase unfeasible.

Yahoo started losing users when Google launched Gmail. Google's UI was far cleaner than Yahoo's.

Yahoo offered $1 billion to buy Facebook in July 2006, but Zuckerberg and the board sought $1.1 billion. Yahoo rejected, and Facebook's valuation rose, making it difficult to buy.

Yahoo was losing users daily while Google and Facebook gained many. Google and Facebook's popularity soared. Yahoo lost value daily.

Microsoft offered $45 billion to buy Yahoo in February 2008, but Yahoo declined. Microsoft increased its bid to $47 billion after Yahoo said it was too low, but Yahoo rejected it. Then Microsoft rejected Yahoo’s 10% bid increase in May 2008.

In 2015, Verizon bought Yahoo for $4.5 billion, and Apollo Global Management bought 90% of Yahoo's shares for $5 billion in May 2021. Verizon kept 10%.

Yahoo's opportunity to acquire Google and Facebook could have been a turning moment. It declined Microsoft's $45 billion deal in 2008 and was sold to Verizon for $4.5 billion in 2015. Poor decisions and lack of vision caused its downfall. Yahoo's aim wasn't obvious and it didn't stick to a single domain.

Hence, a corporation needs a clear vision and a leader who can see its future.

Liked this article? Join my tech and programming newsletter here.

Deep learning-based generative art is being researched. As usual, self-learning is better. Some models, like OpenAI's DALL-E 2, require registration and can only be used online, but others can be used locally, which is usually more enjoyable for curious users. I'll demonstrate the Stable Diffusion model's operation on a standard PC.

Let’s get started.

What It Does

Stable Diffusion uses numerous components:

• A generative model trained to produce images is called a diffusion model. The model is incrementally improving the starting data, which is only random noise. The model has an image, and while it is being trained, the reversed process is being used to add noise to the image. Being able to reverse this procedure and create images from noise is where the true magic is (more details and samples can be found in the paper).

• An internal compressed representation of a latent diffusion model, which may be altered to produce the desired images, is used (more details can be found in the paper). The capacity to fine-tune the generation process is essential because producing pictures at random is not very attractive (as we can see, for instance, in Generative Adversarial Networks).

• A neural network model called CLIP (Contrastive Language-Image Pre-training) is used to translate natural language prompts into vector representations. This model, which was trained on 400,000,000 image-text pairs, enables the transformation of a text prompt into a latent space for the diffusion model in the scenario of stable diffusion (more details in that paper).

This figure shows all data flow:

The weights file size for Stable Diffusion model v1 is 4 GB and v2 is 5 GB, making the model quite huge. The v1 model was trained on 256x256 and 512x512 LAION-5B pictures on a 4,000 GPU cluster using over 150.000 NVIDIA A100 GPU hours. The open-source pre-trained model is helpful for us. And we will.

Install

Before utilizing the Python sources for Stable Diffusion v1 on GitHub, we must install Miniconda (assuming Git and Python are already installed):

wget https://repo.anaconda.com/miniconda/Miniconda3-py39_4.12.0-Linux-x86_64.sh
chmod +x Miniconda3-py39_4.12.0-Linux-x86_64.sh
./Miniconda3-py39_4.12.0-Linux-x86_64.sh
conda update -n base -c defaults conda

Install the source and prepare the environment:

git clone https://github.com/CompVis/stable-diffusion
cd stable-diffusion
conda env create -f environment.yaml
conda activate ldm
pip3 install transformers --upgrade

Download the pre-trained model weights next. HiggingFace has the newest checkpoint sd-v14.ckpt (a download is free but registration is required). Put the file in the project folder and have fun:

python3 scripts/txt2img.py --prompt "hello world" --plms --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

Almost. The installation is complete for happy users of current GPUs with 12 GB or more VRAM. RuntimeError: CUDA out of memory will occur otherwise. Two solutions exist.

Running the optimized version

Try optimizing first. After cloning the repository and enabling the environment (as previously), we can run the command:

python3 optimizedSD/optimized_txt2img.py --prompt "hello world" --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

Stable Diffusion worked on my visual card with 8 GB RAM (alas, I did not behave well enough to get NVIDIA A100 for Christmas, so 8 GB GPU is the maximum I have;).

Running Stable Diffusion without GPU

If the GPU does not have enough RAM or is not CUDA-compatible, running the code on a CPU will be 20x slower but better than nothing. This unauthorized CPU-only branch from GitHub is easiest to obtain. We may easily edit the source code to use the latest version. It's strange that a pull request for that was made six months ago and still hasn't been approved, as the changes are simple. Readers can finish in 5 minutes:

• Replace if attr.device!= torch.device(cuda) with if attr.device!= torch.device(cuda) and torch.cuda.is available at line 20 of ldm/models/diffusion/ddim.py ().

• Replace if attr.device!= torch.device(cuda) with if attr.device!= torch.device(cuda) and torch.cuda.is available in line 20 of ldm/models/diffusion/plms.py ().

• Replace device=cuda in lines 38, 55, 83, and 142 of ldm/modules/encoders/modules.py with device=cuda if torch.cuda.is available(), otherwise cpu.

• Replace model.cuda() in scripts/txt2img.py line 28 and scripts/img2img.py line 43 with if torch.cuda.is available(): model.cuda ().

Run the script again.

Testing

Test the model. Text-to-image is the first choice. Test the command line example again:

python3 scripts/txt2img.py --prompt "hello world" --plms --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

The slow generation takes 10 seconds on a GPU and 10 minutes on a CPU. Final image:

Hello world is dull and abstract. Try a brush-wielding hamster. Why? Because we can, and it's not as insane as Napoleon's cat. Another image:

Generating an image from a text prompt and another image is interesting. I made this picture in two minutes using the image editor (sorry, drawing wasn't my strong suit):

I can create an image from this drawing:

python3 scripts/img2img.py --prompt "A bird is sitting on a tree branch" --ckpt sd-v1-4.ckpt --init-img bird.png --strength 0.8

It was far better than my initial drawing:

I hope readers understand and experiment.

Stable Diffusion UI

Developers love the command line, but regular users may struggle. Stable Diffusion UI projects simplify image generation and installation. Simple usage:

• Unpack the ZIP after downloading it from https://github.com/cmdr2/stable-diffusion-ui/releases. Linux and Windows are compatible with Stable Diffusion UI (sorry for Mac users, but those machines are not well-suitable for heavy machine learning tasks anyway;).

• Start the script.

Done. The web browser UI makes configuring various Stable Diffusion features (upscaling, filtering, etc.) easy:

V2.1 of Stable Diffusion

I noticed the notification about releasing version 2.1 while writing this essay, and it was intriguing to test it. First, compare version 2 to version 1:

• alternative text encoding. The Contrastive LanguageImage Pre-training (CLIP) deep learning model, which was trained on a significant number of text-image pairs, is used in Stable Diffusion 1. The open-source CLIP implementation used in Stable Diffusion 2 is called OpenCLIP. It is difficult to determine whether there have been any technical advancements or if legal concerns were the main focus. However, because the training datasets for the two text encoders were different, the output results from V1 and V2 will differ for the identical text prompts.

• a new depth model that may be used to the output of image-to-image generation.

• a revolutionary upscaling technique that can quadruple the resolution of an image.

• Generally higher resolution Stable Diffusion 2 has the ability to produce both 512x512 and 768x768 pictures.

The Hugging Face website offers a free online demo of Stable Diffusion 2.1 for code testing. The process is the same as for version 1.4. Download a fresh version and activate the environment:

conda deactivate  
conda env remove -n ldm  # Use this if version 1 was previously installed
git clone https://github.com/Stability-AI/stablediffusion
cd stablediffusion
conda env create -f environment.yaml
conda activate ldm

Hugging Face offers a new weights ckpt file.

The Out of memory error prevented me from running this version on my 8 GB GPU. Version 2.1 fails on CPUs with the slow conv2d cpu not implemented for Half error (according to this GitHub issue, the CPU support for this algorithm and data type will not be added). The model can be modified from half to full precision (float16 instead of float32), however it doesn't make sense since v1 runs up to 10 minutes on the CPU and v2.1 should be much slower. The online demo results are visible. The same hamster painting with a brush prompt yielded this result:

It looks different from v1, but it functions and has a higher resolution.

The superresolution.py script can run the 4x Stable Diffusion upscaler locally (the x4-upscaler-ema.ckpt weights file should be in the same folder):

python3 scripts/gradio/superresolution.py configs/stable-diffusion/x4-upscaling.yaml x4-upscaler-ema.ckpt

This code allows the web browser UI to select the image to upscale:

The copy-paste strategy may explain why the upscaler needs a text prompt (and the Hugging Face code snippet does not have any text input as well). I got a GPU out of memory error again, although CUDA can be disabled like v1. However, processing an image for more than two hours is unlikely:

Stable Diffusion Limitations

When we use the model, it's fun to see what it can and can't do. Generative models produce abstract visuals but not photorealistic ones. This fundamentally limits The generative neural network was trained on text and image pairs, but humans have a lot of background knowledge about the world. The neural network model knows nothing. If someone asks me to draw a Chinese text, I can draw something that looks like Chinese but is actually gibberish because I never learnt it. Generative AI does too! Humans can learn new languages, but the Stable Diffusion AI model includes only language and image decoder brain components. For instance, the Stable Diffusion model will pull NO WAR banner-bearers like this:

V1:

V2.1:

The shot shows text, although the model never learned to read or write. The model's string tokenizer automatically converts letters to lowercase before generating the image, so typing NO WAR banner or no war banner is the same.

I can also ask the model to draw a gorgeous woman:

V1:

V2.1:

The first image is gorgeous but physically incorrect. A second one is better, although it has an Uncanny valley feel. BTW, v2 has a lifehack to add a negative prompt and define what we don't want on the image. Readers might try adding horrible anatomy to the gorgeous woman request.

If we ask for a cartoon attractive woman, the results are nice, but accuracy doesn't matter:

V1:

V2.1:

Another example: I ordered a model to sketch a mouse, which looks beautiful but has too many legs, ears, and fingers:

V1:

V2.1: improved but not perfect.

V1 produces a fun cartoon flying mouse if I want something more abstract:

I tried multiple times with V2.1 but only received this:

The image is OK, but the first version is closer to the request.

Stable Diffusion struggles to draw letters, fingers, etc. However, abstract images yield interesting outcomes. A rural landscape with a modern metropolis in the background turned out well:

V1:

V2.1:

Generative models help make paintings too (at least, abstract ones). I searched Google Image Search for modern art painting to see works by real artists, and this was the first image:

I typed "abstract oil painting of people dancing" and got this:

V1:

V2.1:

It's a different style, but I don't think the AI-generated graphics are worse than the human-drawn ones.

The AI model cannot think like humans. It thinks nothing. A stable diffusion model is a billion-parameter matrix trained on millions of text-image pairs. I input "robot is creating a picture with a pen" to create an image for this post. Humans understand requests immediately. I tried Stable Diffusion multiple times and got this:

This great artwork has a pen, robot, and sketch, however it was not asked. Maybe it was because the tokenizer deleted is and a words from a statement, but I tried other requests such robot painting picture with pen without success. It's harder to prompt a model than a person.

I hope Stable Diffusion's general effects are evident. Despite its limitations, it can produce beautiful photographs in some settings. Readers who want to use Stable Diffusion results should be warned. Source code examination demonstrates that Stable Diffusion images feature a concealed watermark (text StableDiffusionV1 and SDV2) encoded using the invisible-watermark Python package. It's not a secret, because the official Stable Diffusion repository's test watermark.py file contains a decoding snippet. The put watermark line in the txt2img.py source code can be removed if desired. I didn't discover this watermark on photographs made by the online Hugging Face demo. Maybe I did something incorrectly (but maybe they are just not using the txt2img script on their backend at all).

Conclusion

The Stable Diffusion model was fascinating. As I mentioned before, trying something yourself is always better than taking someone else's word, so I encourage readers to do the same (including this article as well;).

Is Generative AI a game-changer? My humble experience tells me:

• I think that place has a lot of potential. For designers and artists, generative AI can be a truly useful and innovative tool. Unfortunately, it can also pose a threat to some of them since if users can enter a text field to obtain a picture or a website logo in a matter of clicks, why would they pay more to a different party? Is it possible right now? unquestionably not yet. Images still have a very poor quality and are erroneous in minute details. And after viewing the image of the stunning woman above, models and fashion photographers may also unwind because it is highly unlikely that AI will replace them in the upcoming years.

• Today, generative AI is still in its infancy. Even 768x768 images are considered to be of a high resolution when using neural networks, which are computationally highly expensive. There isn't an AI model that can generate high-resolution photographs natively without upscaling or other methods, at least not as of the time this article was written, but it will happen eventually.

• It is still a challenge to accurately represent knowledge in neural networks (information like how many legs a cat has or the year Napoleon was born). Consequently, AI models struggle to create photorealistic photos, at least where little details are important (on the other side, when I searched Google for modern art paintings, the results are often even worse;).

• When compared to the carefully chosen images from official web pages or YouTube reviews, the average output quality of a Stable Diffusion generation process is actually less attractive because to its high degree of randomness. When using the same technique on their own, consumers will theoretically only view those images as 1% of the results.

Anyway, it's exciting to witness this area's advancement, especially because the project is open source. Google's Imagen and DALL-E 2 can also produce remarkable findings. It will be interesting to see how they progress.