One Sided Gaussian Filter

Notes:

These indicators and concepts are specifically designed for TradingView.com

Overview

One-sided Gaussian filters are a type of technical indicator used in trading to smooth out price data and identify trends. The filter is based on the Gaussian distribution, which is a statistical distribution that represents the probability of a variable taking a certain value. The filter is designed to place more weight on recent price data than on older data, which can help traders identify trends and potential entry and exit points.

How to Trade

This indicator will print "L" for long and "S" for short which can be traded. It can give false "S" or "L" so only take them if the price breaks out of the white dashed bands indicating a very strong trend in that direction. There are also a variety of ways to use it differently:

1. Trend identification: Traders can use the filter to identify the direction of the trend. If the filtered price is rising, it suggests a bullish trend, while a falling filtered price suggests a bearish trend.
2. Entry and exit points: Traders can use the filter to identify potential entry and exit points. When the filtered price crosses above the current price, it can be a buy signal, while a cross below can be a sell signal.
3. Stop loss placement: Traders can use the filter to determine where to place their stop loss orders. The filter can help identify areas where the price is likely to turn around, which can be used to set stop loss orders at appropriate levels.

How to Calculate

Here is how to calculate a one-sided Gaussian filter:

1. Choose a period for the filter. This is typically based on the trader's preference and the timeframe of the chart being analyzed.
2. Determine the standard deviation for the filter. This is also based on the trader's preference and can be adjusted to make the filter more or less sensitive to price changes.
3. Calculate the weight for each price data point using the Gaussian distribution. The weight is calculated using the following formula:

Weight = e ^ (-0.5 * ((n-i)/sigma) ^ 2)

Where:

• e = the mathematical constant 2.71828...
• n = the period of the filter
• i = the current data point being analyzed
• sigma = the standard deviation of the filter

4. Multiply each price data point by its weight to get the filtered price.