Skip to content

Moving Averages

Pypulate provides a comprehensive set of moving average functions for financial time series analysis. This page explains the different types of moving averages available and how to use them.

Available Moving Averages

Specialized Moving Averages

These moving averages are designed for specific use cases:

  • Volume-Weighted Moving Average (VWMA): Weights price by volume.
  • Kaufman Adaptive Moving Average (KAMA): Adapts to market volatility.
  • Arnaud Legoux Moving Average (ALMA): Reduces lag and noise.
  • Fractal Adaptive Moving Average (FRAMA): Adapts to market fractal dimension.
  • Jurik Moving Average (JMA): Reduces noise and lag.
  • Laguerre Filter: Uses Laguerre polynomials for smoothing.
  • Least Squares Moving Average (LSMA): Uses linear regression.
  • McGinley Dynamic Indicator: Adapts to market speed.
  • Modular Filter: Adjusts smoothing based on phase.
  • Rex Dog Moving Average (RDMA): Average of six SMAs with different periods.
  • Tillson T3: Triple EMA with reduced lag.
  • Volatility-Adjusted Moving Average (VAMA): Adjusts based on volatility.

Using Moving Averages

Functional Approach

You can use moving averages directly by importing the functions:

import numpy as np
from pypulate.moving_averages import sma, ema, hma

# Create sample data
data = [10, 11, 12, 11, 10, 9, 10, 11, 12, 13]

# Calculate moving averages
sma_result = sma(data, period=3)
ema_result = ema(data, period=3)
hma_result = hma(data, period=3)

Method Chaining with Parray

For a more fluent interface, you can use the Parray class:

from pypulate import Parray

# Create sample data
data = Parray([10, 11, 12, 11, 10, 9, 10, 11, 12, 13])

# Calculate moving averages using method chaining
sma_result = Parray.sma(period=3)
ema_result = Parray.ema(period=3)
hma_result = Parray.hma(period=3)

Examples

Comparing Different Moving Averages

import numpy as np
import matplotlib.pyplot as plt
from pypulate import Parray

# Generate sample price data
np.random.seed(42)
days = 100
price = np.cumsum(np.random.normal(0, 1, days)) + 100

# Convert to Parray for method chaining
price_array = Parray(price)

# Calculate different types of moving averages
sma = price_array.sma(20)
ema = price_array.ema(20)
wma = price_array.wma(20)
hma = price_array.hma(20)

# Plot the results
plt.figure(figsize=(12, 6))
plt.plot(price, label='Price', alpha=0.5, color='gray')
plt.plot(sma, label='SMA(20)')
plt.plot(ema, label='EMA(20)')
plt.plot(wma, label='WMA(20)')
plt.plot(hma, label='HMA(20)')

plt.title('Comparison of Different Moving Averages')
plt.xlabel('Days')
plt.ylabel('Price')
plt.legend()
plt.grid(True, alpha=0.3)
plt.show()

Moving Average Crossover Strategy

import numpy as np
import matplotlib.pyplot as plt
from pypulate import Parray

# Generate sample price data
np.random.seed(42)
days = 200
price = np.cumsum(np.random.normal(0, 1, days)) + 100

# Convert to Parray for method chaining
price_array = Parray(price)

# Calculate fast and slow EMAs
fast_ema = price_array.ema(9)
slow_ema = price_array.ema(21)

# Generate buy/sell signals
buy_signals = fast_ema.crossover(slow_ema)
sell_signals = fast_ema.crossunder(slow_ema)

# Plot the results
plt.figure(figsize=(12, 6))
plt.plot(price, label='Price')
plt.plot(fast_ema, label='9-day EMA', alpha=0.7)
plt.plot(slow_ema, label='21-day EMA', alpha=0.7)

# Plot buy signals
buy_indices = np.where(buy_signals)[0]
plt.scatter(buy_indices, price[buy_indices], marker='^', color='green', s=100, label='Buy Signal')

# Plot sell signals
sell_indices = np.where(sell_signals)[0]
plt.scatter(sell_indices, price[sell_indices], marker='v', color='red', s=100, label='Sell Signal')

plt.title('Moving Average Crossover Strategy')
plt.xlabel('Days')
plt.ylabel('Price')
plt.legend()
plt.grid(True, alpha=0.3)
plt.show()

Choosing the Right Moving Average

Different moving averages are suitable for different market conditions:

  • Trending Markets: SMA, EMA, WMA, HMA
  • Volatile Markets: KAMA, ALMA, FRAMA, JMA
  • Ranging Markets: DEMA, TEMA, T3

Experiment with different types to find the one that works best for your specific use case.