Chapter 35 Technical Analysis for Stocks using Plotly
Ivan Ugalde (du2160); Bernardo Lopez (bl2786)
The goal of this R markdown file is to show how to use Apha Vantage API to download stocks data, organize it and make interactive plots using Plotly that will help to find trends in the market.
35.1 Import all libraries
We will be ussing:
- plotly: For interactive plots.
- alphavantager: R package from Alpha Vantage for downloading stocks data.
- tidyverse
35.2 Download data from Alpha Vantage
Alpha Vantage provides free historical and real time information on stocks via an API.
35.2.1 Usefull links for more information:
Alpha Vantage API: alphavantager R package documentation.
RSI: Relative strength index.
MACD: Moving Average Convergence Divergence.
PSAR: Parabolic SAR (Stop and Reverse) indicator.
We will store in symbol the code for the stock we want to analyse. In this example we wil use Apple (AAPL), but you can change it to whatever you want. Manchester United (MANU), Starbucks (SBUX), Netflix (NFLX) and New York Times (NYT) are some examples. In start_date we store the date from which we want to start analyzing the stock.
With the API we will download data for the selected stock from 2018-01-01 till today. We will download the daily value of the stock and the indicators SAR, MACD and RSI.
In some cases the API returns the date also with the hour, so we will use the function mutate and as.Date() to have the same date format in all the API calls.
The function av_get from the package alphavantager is used to get financial data from the Alpha Vantage API. It requires two main parameters: symbol for the code of the stock and av_fun, the Alpha Vantage function that describes the type of data you want to get. For this example we will use “TIME_SERIES_DAILY” for the daily value of the stock, and “SAR”, “MACD” and “RSI” to get those financial indicators.
Tu use Alpha Vantage API you have to create your own key, it is free. To create the html file we used our key, but you will have to replace your new key in the chunk bellow.
av_api_key(api_key)
stock_symbol <- 'AAPL'
start_date <- as.Date('2018-01-01')
ohlc <- av_get(symbol = stock_symbol, av_fun = 'TIME_SERIES_DAILY', outputsize = 'full') %>%
filter(timestamp >= start_date)
psar <- av_get(symbol = stock_symbol, av_fun = 'SAR', interval='daily') %>%
filter(time >= start_date) %>%
mutate(time=as.Date(time,format="%Y-%m-%d"))
macd <- av_get(symbol = stock_symbol, av_fun = 'MACD', interval='daily', series_type = 'close') %>%
filter(time >= start_date) %>%
mutate(time=as.Date(time,format="%Y-%m-%d"))
rsi <- av_get(symbol = stock_symbol, av_fun = 'RSI', interval='daily', time_period=14, series_type='close') %>%
filter(time >= start_date) %>%
mutate(time=as.Date(time,format="%Y-%m-%d"))The following code merges all of the data using the date as binder.
data <- ohlc %>%
merge(y = psar, by.x = 'timestamp', by.y = 'time') %>%
merge(y = macd, by.x = 'timestamp', by.y = 'time') %>%
merge(y = rsi, by.x = 'timestamp', by.y = 'time')
data %>% head()## timestamp open high low close volume sar macd macd_hist
## 1 2018-01-02 170.16 172.30 169.26 172.26 25555900 167.3888 0.3870 -0.4921
## 2 2018-01-03 172.53 174.55 171.96 172.23 29517900 167.4869 0.3751 -0.4032
## 3 2018-01-04 172.54 173.47 172.08 173.03 22434600 167.5841 0.4254 -0.2823
## 4 2018-01-05 173.44 175.37 173.05 175.00 23660000 167.6802 0.6170 -0.0725
## 5 2018-01-08 174.35 175.61 173.93 174.35 20567800 167.7754 0.7083 0.0150
## 6 2018-01-09 174.55 175.06 173.41 174.33 21584000 167.8697 0.7702 0.0615
## macd_signal rsi
## 1 0.8791 51.9084
## 2 0.7783 51.8233
## 3 0.7077 53.9903
## 4 0.6896 58.8936
## 5 0.6933 56.7448
## 6 0.7087 56.676335.3 Simple plot: 2 traces in same axis
To add traces to the plot we use add_trace. In this example we will use as type “candlestick” and “scatter”.
plot1 <- plot_ly(data) %>%
add_trace(type = 'candlestick',
name = 'OHLC',
x = ~timestamp,
open = ~open, high = ~high, low = ~low, close = ~close,
increasing = list(line = list(color='rgba(52,169,102,1)',
width=1),
fillcolor = 'rgba(0,0,0,0)'), # Transparent
decreasing = list(line = list(color='rgba(220,68,59,1)',
width=1),
fillcolor = 'rgba(0,0,0,0)'), # Transparent
legendgroup = 'one') %>%
add_trace(type = 'scatter',
mode = 'markers',
x = ~timestamp,
y = ~sar,
name = 'PSAR',
marker = list(color = 'orange', size = 4),
legendgroup = 'one')
plot135.4 Many traces in independent axis but in same plot
Plotly give us interactivity, that allows to zoom the data for the same dates for all graphs simultanously.
By using the parameter yaxis we can plot traces in different Y axis but same X axis.
plot2 <- plot1 %>%
add_trace(type = 'bar',
x = ~timestamp,
y = ~macd_hist,
name = 'MACD Histogram',
marker = list(color = 'gray'),
yaxis = 'y2',
legendgroup = 'two') %>%
add_trace(type = 'scatter',
mode = 'lines',
marker = NULL,
x = ~timestamp,
y = ~macd,
name = 'MACD',
line = list(color = 'red'),
yaxis = 'y2',
legendgroup = 'two') %>%
add_trace(type = 'scatter',
mode = 'lines',
marker = NULL,
x = ~timestamp,
y = ~macd_signal,
name = 'Signal',
line = list(color = 'plum'),
yaxis = 'y2',
legendgroup = 'two') %>%
add_trace(type = 'scatter',
mode = 'lines',
marker = NULL,
x = ~timestamp,
y = ~rsi,
name = 'RSI',
line = list(color = 'plum'),
yaxis = 'y3',
legendgroup = 'three') %>%
add_trace(type = 'scatter',
mode = 'lines',
marker = NULL,
x = c(~min(timestamp), ~max(timestamp)),
y = c(70,70),
name = 'RSI',
line = list(color = 'red',
width = 0.5,
dash = 'dot'),
yaxis = 'y3',
legendgroup = 'three') %>%
add_trace(type = 'scatter',
mode = 'lines',
marker = NULL,
x = c(~min(timestamp), ~max(timestamp)),
y = c(30,30),
name = 'RSI',
line = list(color = 'red',
width = 0.5,
dash = 'dot'),
yaxis = 'y3',
legendgroup = 'three') %>%
layout(yaxis = list(domain = c(0.62, 1),
fixedrange = FALSE),
yaxis2 = list(domain = c(0.32, 0.58),
fixedrange = FALSE),
yaxis3 = list(domain = c(0., 0.28),
fixedrange = FALSE),
height = 500)
plot235.5 Aesthetics: background and margins
We use parameter paper_bgcolor and plot_bgcolor of layout to change the color of the background of the paper and the plot, we decided to use black. Parameter margin is used to change the size of the plot margins.
35.6 More aesthetics: hide legends and hide X-axis slider
To hide the legend we must set the parameter showlegend as false.
For every Y-axis we used parameters titlte to set the title of the axis and titlefont and tickfont to change the colors of the title and the ticks.
plot4 <- plot3 %>%
layout(xaxis = list(titlefont = list(color='rgb(200,115,115)'),
tickfont = list(color='rgb(200,200,200)'),
linewidth=1,
linecolor = 'white'),
yaxis = list(domain = c(0.62, 1),
title = 'PSAR & OHLC',
titlefont = list(color='rgb(200,115,115)'),
tickfont = list(color='rgb(200,200,200)'),
linewidth=1,
linecolor = 'white',
mirror = "all"),
yaxis2 = list(domain = c(0.32, 0.58),
title = 'MACD',
titlefont = list(color='rgb(200,115,115)'),
tickfont = list(color='rgb(200,200,200)'),
linewidth=1,
linecolor = 'white',
mirror = "all"),
yaxis3 = list(domain = c(0., 0.28),
title = 'RSI',
titlefont = list(color='rgb(200,115,115)'),
tickfont = list(color='rgb(200,200,200)'),
linewidth=1,
linecolor = 'white',
mirror = "all"),
showlegend = FALSE,
height = 500)
plot435.7 Shortcuts to slice data by pre-fixed date ranges
This last chunk of code will add a rangeselector to the x axis. This buttons will allow us to focus only in the last 3, 6 or 12 months. We can also see all the data or see it year to date (from the begining of the year to the current day). For each button in the rangeselector we have to specify the count, label, step and stepmode.
plot5 <- plot4 %>%
layout(xaxis = list( rangeselector = list( buttons = list( list(count = 3,
label = "3 mo",
step = "month",
stepmode = "backward"),
list(
count = 6,
label = "6 mo",
step = "month",
stepmode = "backward"),
list(
count = 1,
label = "1 yr",
step = "year",
stepmode = "backward"),
list(
count = 1,
label = "YTD",
step = "year",
stepmode = "todate"),
list(step = "all"))),
rangeslider=list(visible=FALSE)))
plot5