predict_bitcoin_price_with_arima_models.R

> library(forecast)
> train = read.csv("c.csv")
> T<-train[which(train$name == 'btc'), ]
> Train<-T[ with(T, order(ts)),]
> head(Train)
#    name           ts usd_value      volume
# 1   btc 1.558271e+12    7912.8 23649292230
# 15  btc 1.558272e+12    7955.9 23741378790
# 29  btc 1.558272e+12    7940.0 23816540813
# 43  btc 1.558272e+12    7942.9 23891912844
# 57  btc 1.558273e+12    7964.6 23914583777
# 71  btc 1.558273e+12    7947.0 23941990903
> holt(Train[,'usd_value'])
#      Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
# 1670       7942.866 7918.622 7967.110 7905.788 7979.943
# 1671       7942.887 7909.384 7976.390 7891.648 7994.126
# 1672       7942.908 7902.198 7983.618 7880.648 8005.169
# 1673       7942.930 7896.108 7989.751 7871.323 8014.536
# 1674       7942.951 7890.728 7995.174 7863.082 8022.819
# 1675       7942.972 7885.854 8000.090 7855.618 8030.326
# 1676       7942.993 7881.368 8004.619 7848.745 8037.241
# 1677       7943.014 7877.188 8008.841 7842.342 8043.687
# 1678       7943.036 7873.261 8012.811 7836.324 8049.747
# 1679       7943.057 7869.544 8016.570 7830.628 8055.485
> thetaf(Train[,'usd_value'])
#      Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
# 1670       7942.862 7918.661 7967.063 7905.850 7979.874
# 1671       7942.887 7909.402 7976.372 7891.677 7994.098
# 1672       7942.913 7902.209 7983.617 7880.662 8005.164
# 1673       7942.938 7896.116 7989.761 7871.329 8014.548
# 1674       7942.964 7890.734 7995.194 7863.086 8022.843
# 1675       7942.990 7885.862 8000.117 7855.621 8030.358
# 1676       7943.015 7881.378 8004.652 7848.750 8037.280
# 1677       7943.041 7877.203 8008.879 7842.350 8043.731
# 1678       7943.066 7873.279 8012.853 7836.336 8049.796
# 1679       7943.092 7869.568 8016.616 7830.647 8055.537
# 
> forecast(Train[,'usd_value'])
#      Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
# 1670       7942.836 7918.636 7967.037 7905.825 7979.848
# 1671       7942.836 7909.358 7976.315 7891.636 7994.037
# 1672       7942.836 7902.144 7983.529 7880.602 8005.071
# 1673       7942.836 7896.028 7989.645 7871.250 8014.423
# 1674       7942.836 7890.624 7995.049 7862.985 8022.688
# 1675       7942.836 7885.729 7999.944 7855.499 8030.174
# 1676       7942.836 7881.222 8004.451 7848.605 8037.068
# 1677       7942.836 7877.023 8008.650 7842.183 8043.490
# 1678       7942.836 7873.076 8012.597 7836.147 8049.526
# 1679       7942.836 7869.340 8016.332 7830.434 8055.239
# 
> auto.arima(Train[,'usd_value'])
# Series: Train[, "usd_value"]
# ARIMA(2,1,3)
#   
# Coefficients:
#           ar1      ar2     ma1     ma2      ma3
#       -1.2890  -0.9429  1.2548  0.8634  -0.0667
# s.e.   0.0347   0.0314  0.0416  0.0519   0.0254
#   
# sigma^2 estimated as 353:  log likelihood=-7256.91
# AIC=14525.82   AICc=14525.87   BIC=14558.33
# 
> fit = auto.arima(Train[,'usd_value'])
> forecast(fit)
#      Point Forecast    Lo 80    Hi 80    Lo 95    Hi 95
# 1670       7943.523 7919.446 7967.601 7906.700 7980.346
# 1671       7942.970 7909.496 7976.444 7891.776 7994.164
# 1672       7943.099 7902.820 7983.379 7881.498 8004.701
# 1673       7943.454 7897.233 7989.675 7872.766 8014.143
# 1674       7942.875 7891.190 7994.560 7863.830 8021.920
# 1675       7943.287 7887.002 7999.572 7857.207 8029.367
# 1676       7943.302 7882.522 8004.082 7850.346 8036.258
# 1677       7942.894 7877.946 8007.842 7843.564 8042.224
# 1678       7943.406 7874.724 8012.088 7838.366 8048.446
# 1679       7943.131 7870.669 8015.593 7832.310 8053.952