0

I have time series data for top 10 cryptocurrencies including these features:

open, close, high, low, volume.

1). I want to know which method is better for correlation testing between them - Pearson's Correlation or Spearman Correlation? Why?

2). Correlation should be tested for which price - low, high, open, close, average?

3). Since it is time series data, for testing correlation how much time period data is required(minimum)?

The_Different
  • 1
  • 2
  • Hi this looks more like an Project. For the Time-series you can use influxdb https://www.influxdata.com for the analyse python pandas –  Dec 23 '21 at 19:53
  • Possibly related (because you appear to be asking about corr( price_it, price_jt ) for two different financial assets i and j): https://stats.stackexchange.com/questions/555540/does-it-really-matter-if-a-correlation-is-spurious – Adrian Dec 23 '21 at 20:09
  • Also related: https://stats.stackexchange.com/questions/133155/how-to-use-pearson-correlation-correctly-with-time-series – Adrian Dec 23 '21 at 20:11
  • @Adrian I already read those solutions but here I am asking in context with cryptocurrencies means they don't follow the assumptions of pearson correlation (normaility). Also please could you answer other two questions as well. – The_Different Dec 24 '21 at 04:06

0 Answers0