OpenBB/website/content/sdk/guides/intros/forecast.md

title, keywords, excerpt

title	keywords	excerpt
Forecast	machine, learning, prediction, forecasting, neural, network, linear, regression, time, series, scripts, data, mining, cleaning, transformer, analyst, equity, research, api, sdk, application, python, notebook, jupyter	This guide introduces the Forecast SDK in the context of the OpenBB SDK.

The Forecast module provides programmatic access to the commands from within the OpenBB Terminal. Import the OpenBB SDK module, and then access the functions similarly to how the Terminal menus are navigated. The code completion will be activated upon entering ., after, openbb.forecast ​

How to Use

​ The examples provided below will assume that the following import block is included at the beginning of the Python script or Notebook file: ​

from openbb_terminal.sdk import openbb
import pandas as pd
%matplotlib inline

​ A brief description below highlights the main Functions and Modules available in the ETF SDK

Path	Type	Description
openbb.forecast.load	Function	Show the data selected data on a chart
openbb.forecast.plot	Function	Plot a specific column of a loaded dataset
openbb.forecast.clean	Function	Clean a dataset by filling or dropping NaNs
openbb.forecast.combine	Function	Combine columns from different datasets
openbb.forecast.desc	Function	Show descriptive statistics of a dataset
openbb.forecast.corr	Function	Plot the correlation coefficients for dataset features
openbb.forecast.season	Function	Plot the seasonality for a dataset column
openbb.forecast.delete	Function	Delete columns from dataset
openbb.forecast.rename	Function	Rename columns from dataset
openbb.forecast.export	Function	Export a processed dataset
openbb.forecast.signal	Function	Add Price Signal (short vs. long term)
openbb.forecast.atr	Function	Add Average True Range
openbb.forecast.ema	Function	Add Exponentially Weighted Moving Average
openbb.forecast.sto	Function	Add Stochastic Oscillator %K and %D
openbb.forecast.rsi	Function	Add Relative Strength Index
openbb.forecast.roc	Function	Add Rate of Change
openbb.forecast.mom	Function	Add Momentum
openbb.forecast.delta	Function	Add % Change
openbb.forecast.autoces	Function	Automatic Complex Exponential Smoothing Model
openbb.forecast.autoets	Function	Automatic ETS (Error, Trend, Seasonality) Model
openbb.forecast.seasonalnaive	Function	Seasonal Naive Model
openbb.forecast.expo	Function	Probabilistic Exponential Smoothing
openbb.forecast.theta	Function	Theta Method
openbb.forecast.linregr	Function	Probabilistic Linear Regression (feat. Past covariates and Explainability)
openbb.forecast.regr	Function	Regression (feat. Past covariates and Explainability)
openbb.forecast.rnn	Function	Probabilistic Recurrent Neural Network (RNN, LSTM, GRU)
openbb.forecast.brnn	Function	Block Recurrent Neural Network (RNN, LSTM, GRU) (feat. Past covariates)
openbb.forecast.nbeats	Function	Neural Bayesian Estimation (feat. Past covariates)
openbb.forecast.tcn	Function	Temporal Convolutional Neural Network (feat. Past covariates)
openbb.forecast.trans	Function	Transformer Network (feat. Past covariates)
openbb.forecast.tft	Function	Temporal Fusion Transformer Network(feat. Past covariates)
openbb.forecast.nhits	Function	Neural Hierarchical Interpolation (feat. Past covariates)
openbb.forecast.rwd	Function	Random Walk with Drift Model
openbb.forecast.autoselect	Function	Select best statistical model from AutoARIMA, AutoETS, AutoCES, MSTL, etc.

Alteratively you can print the contents of the Forecast SDK with: ​

help(openbb.forecast)

Examples

corr

​ Shows a correlation matrix between columns in dataset ​

df = pd.read_csv(ANDREW_REPLACE)
openbb.forecast.corr_chart(df)

​

season

​ This command allows you to see seasonality patterns in your data

df = pd.read_csv(ANDREW_REPLACE)
openbb.forecast.season_chart(df)

​

​

expo

Predicts the future value of time series data using exponential smoothing ​

df = pd.read_csv(ANDREW_REPLACE)
openbb.forecast.expo_chart(df)

​ Here was can now see a chart and table with the expected values, and historic data. The chart also tells us how our backtesing performed, so that we can know the accuracy of our prediction. ​

​