Hi there. In this post, I transfer crypto fear and greed index data into R. The website where I extract the data is from is here.
The crypto fear and greed index is one indicator which gauges to sentiment of people in the crypto market. People in the crypto market can be very emotional. Greed tends to occur when the market is rising and people get FOMO (Fear of Missing Out). Fear can occur when people panic and sell when there is bad news and sharp downturns in price.
Extreme fear in the market can be good buying opportunities.
Extreme greed in the market could mean that the market is due for a correction (drop in price).
Reference: https://alternative.me/crypto/fear-and-greed-index/#api
Start with loading these libraries into R. Use install.packages('pkg_name') if you need to install a certain package into R.
# Load packages, use install.packages("rjson") to install rjson
library(httr)
library(plotly)
library(jsonlite)
# Reference: https://alternative.me/crypto/fear-and-greed-index/#api
url <- "https://api.alternative.me/fng/?limit=0"
# Obtaining fear and greed crypto data
response <- httr::GET(url)
raw_content <- httr::content(response)
fear_greed_data <- jsonlite::fromJSON(rawToChar(response$content))
fear_greed_df <- fear_greed_data$data
# Check fear and greed data with head():
head(fear_greed_df)## value value_classification timestamp time_until_update
## 1 71 Greed 1630022400 18716
## 2 75 Greed 1629936000 <NA>
## 3 73 Greed 1629849600 <NA>
## 4 79 Extreme Greed 1629763200 <NA>
## 5 79 Extreme Greed 1629676800 <NA>
## 6 76 Extreme Greed 1629590400 <NA>
Changing Data Types
The fear and greed index values are in a character format. Timestamps are in a character format as well. Both of these need to be converted into numeric values.
The numeric timestamps need to be converted into a date format. This is done with combining as.POSIXct() with as.Date().
# Convert into numeric for value and timestamp columns
fear_greed_df$value <- as.numeric(fear_greed_df$value)
fear_greed_df$timestamp <- as.numeric(fear_greed_df$timestamp)
# Convert timestamp into date format:
# Example: as.Date(as.POSIXct(1629763200, origin = "1970-01-01"))
fear_greed_df$timestamp <- as.Date(as.POSIXct(fear_greed_df$timestamp, origin = "1970-01-01"))
# Check dataframe:
head(fear_greed_df)## value value_classification timestamp time_until_update
## 1 71 Greed 2021-08-27 18716
## 2 75 Greed 2021-08-26 <NA>
## 3 73 Greed 2021-08-25 <NA>
## 4 79 Extreme Greed 2021-08-24 <NA>
## 5 79 Extreme Greed 2021-08-23 <NA>
## 6 76 Extreme Greed 2021-08-22 <NA>str(fear_greed_df)## 'data.frame': 1301 obs. of 4 variables:
## $ value : num 71 75 73 79 79 76 78 70 70 73 ...
## $ value_classification: chr "Greed" "Greed" "Greed" "Extreme Greed" ...
## $ timestamp : Date, format: "2021-08-27" "2021-08-26" ...
## $ time_until_update : chr "18716" NA NA NA ...
###### Line Graph Plotting:
fear_greed_df %>%
plot_ly(x = ~timestamp, y = ~value, type = 'scatter', mode = 'lines',
width = 900, height = 700) %>%
layout(title = "Crypto Fear & Greed Index",
xaxis = list(title = "\n Date"),
yaxis = list(title = "Index Value \n"))
Creating the bar graph for fear & greed indicator counts is actually a bit involved. The table() function in R is used to obtain counts for each fear and greed indicator
fg_counts <- data.frame(table(fear_greed_df$value_classification))
fg_counts## Var1 Freq
## 1 Extreme Fear 302
## 2 Extreme Greed 171
## 3 Fear 450
## 4 Greed 261
## 5 Neutral 117
Next, I rename the columns in the fg_counts dataframe. The factor() function is for ordering the fear and greed indicators. I have this order.
# Change column names
colnames(fg_counts) <- c('Label', 'Count')
# Reorder Categories (Extreme Fear to Extreme Greed):
fg_counts$Label <- factor(fg_counts$Label,
label = c("Extreme Fear", "Fear", "Neutral", "Greed", "Extreme Greed"))
The dataframe is ready to be used for the bar graph in plotly. I have added the count values as text at the top of the bars with text = ~Count and textposition = 'auto'.
Individual colours for the bars were included to this bargraph. This is done with the use of marker = with a list of colours. The colours are in the RGB format.
# From Feb 1, 2018 to Aug 27, 2021
fg_bargraph <- fg_counts %>%
plot_ly(x = ~Label, y = ~Count, width = 900, height = 500,
name = "Fear & Greed Index Counts \n Since Feb 1, 2018", type = "bar",
text = ~Count, textposition = 'auto',
marker = list(color = c('rgba(222,45,38,0.8)', 'rgba(206, 140, 104, 1)',
'rgba(216, 211, 105, 1)', 'rgba(178, 212, 122, 1)',
'rgba(83, 124, 17, 1)'))) %>%
layout(title = "Fear & Greed Index Day Counts \n Since Feb 1, 2018",
xaxis = list(title = "\n Label"),
yaxis = list(title = "\n Number of Days"))
fg_bargraph
Developing the pie chart in R with plotly is not too difficult. Do specify the type as pie with type = 'pie'. To obtain percentages and counts make sure to have textinfo = 'label+percent+value'. From the documentation https://plotly.com/python/pie-charts/ they do not have this mentioned. I got lucky and guessed it right with textinfo = 'label+percent+value'.
#### Pie Chart
# label+percent+value to obtain labels, counts and percents
fg_piechart <- fg_counts %>%
plot_ly(name = "Fear & Greed Index From Feb 1, 2018 To Aug 27, 2021", type = 'pie',
labels = ~Label, values = ~Count,
width = 900, height = 500,
textinfo='label+percent+value',
insidetextorientation='radial') %>%
layout(title = "Fear & Greed Index Day Counts Feb 1, 2018 To Aug 27, 2021 \n")
fg_piechart