OLS Regression Results
==============================================================================
Dep. Variable: BTCUSD R-squared: 0.989
Model: OLS Adj. R-squared: 0.985
Method: Least Squares F-statistic: 260.6
Date: Sun, 22 Apr 2018 Prob (F-statistic): 1.87e-171
Time: 13:10:27 Log-Likelihood: -2119.3
No. Observations: 280 AIC: 4383.
Df Residuals: 208 BIC: 4644.
Df Model: 71
Covariance Type: nonrobust
==========================================================================================================
coef std err t P>|t| [0.025 0.975]
----------------------------------------------------------------------------------------------------------
Intercept -3.013e+05 1.8e+05 -1.674 0.096 -6.56e+05 5.36e+04
howtobuycryptocurrencyWorldwide[T.1] 284.2228 436.490 0.651 0.516 -576.289 1144.735
howtobuycryptocurrencyWorldwide[T.2] -834.5288 918.605 -0.908 0.365 -2645.499 976.442
howtobuycryptocurrencyWorldwide[T.3] -1639.0373 892.061 -1.837 0.068 -3397.677 119.603
howtobuycryptocurrencyWorldwide[T.4] -1822.9216 1349.968 -1.350 0.178 -4484.296 838.453
howtobuycryptocurrencyWorldwide[T.5] -461.3566 751.629 -0.614 0.540 -1943.144 1020.431
howtobuycryptocurrencyWorldwide[T.6] -1590.4795 1084.831 -1.466 0.144 -3729.153 548.194
howtobuycryptocurrencyWorldwide[T.7] -667.8484 506.288 -1.319 0.189 -1665.962 330.265
howtobuycryptocurrencyWorldwide[T.8] -575.7590 1297.502 -0.444 0.658 -3133.698 1982.180
howtobuycryptocurrencyWorldwide[T.9] -2449.3509 1565.416 -1.565 0.119 -5535.466 636.764
howtobuycryptocurrencyWorldwide[T.10] 1362.5353 1131.645 1.204 0.230 -868.429 3593.499
howtobuycryptocurrencyWorldwide[T.11] 1.206e+04 5006.070 2.408 0.017 2186.460 2.19e+04
howtobuycryptocurrencyWorldwide[T.13] -8135.2934 3056.663 -2.661 0.008 -1.42e+04 -2109.283
howtobuycryptocurrencyWorldwide[T.14] -333.8614 1012.361 -0.330 0.742 -2329.665 1661.943
howtobuycryptocurrencyWorldwide[T.17] -9448.2497 3586.911 -2.634 0.009 -1.65e+04 -2376.888
howtobuycryptocurrencyWorldwide[T.19] -8515.1383 3795.035 -2.244 0.026 -1.6e+04 -1033.475
howtobuycryptocurrencyWorldwide[T.35] -4.1140 1172.341 -0.004 0.997 -2315.308 2307.080
howtobuycryptocurrencyWorldwide[T.36] -1.713e+04 6089.825 -2.814 0.005 -2.91e+04 -5128.168
howtobuycryptocurrencyWorldwide[T.54] -1.193e+04 4885.490 -2.441 0.015 -2.16e+04 -2294.187
howtobuycryptocurrencyWorldwide[T.62] -1.653e+04 5836.682 -2.833 0.005 -2.8e+04 -5027.678
howtobuycryptocurrencyWorldwide[T.72] -1.193e+04 4509.585 -2.645 0.009 -2.08e+04 -3038.531
howtobuycryptocurrencyWorldwide[T.95] -8206.0353 3263.856 -2.514 0.013 -1.46e+04 -1771.556
howtobuycryptocurrencyWorldwide[T.100] -2.327e+04 8503.289 -2.737 0.007 -4e+04 -6507.457
howtobuycryptocurrencyWorldwide[T.<1] -72.6343 359.855 -0.202 0.840 -782.065 636.797
运行线性回归的Python代码：
mod3 = smf.ols('BTCUSD ~ <other variables> +howtobuycryptocurrencyWorldwide+howtobuybitcoinWorldwide+bitcoinWorldwide+howtobuyethereumWorldwide+ethereumWorldwide+howtobuyrippleWorldwide+rippleWorldwide+howtobuylitecoinWorldwide+litecoinWorldwide+bitcoinWorldwideYoutube+ethereumWorldwideYoutube+rippleWorldwideYoutube+litecoinWorldwideYoutube+vitalikWorldwide+satoshiWorldwide',data=cryptos).fit()
print(mod3.summary())
我不理解预测变量[T.x]符号。有人可以解释一下吗？
答案 0 :(得分：0)
问题是Google趋势数据已经＆＃39;＆lt; 1＆＃39;在必须转换的结果中。 我基本上在密码是Dataframe的地方。
cryptos.replace('<1', 0.1 , inplace=True)