R Notebook

Summary of Gas Price for January Observations

jan <- read.csv("January_Data.csv")
rid <- c(1,2,3)
jan <- jan[,-rid]
summary(jan$gasprice)

    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
   0.018   65.000   92.000  121.009  128.000 4379.700 

Simple linear Regresion on scaled gas price and past median

janscale <- scale(jan)
janscale <- as.data.frame(jan)
janmodel <- lm(jan$gasprice~jan$past_median)
summary(janmodel)

Call:
lm(formula = jan$gasprice ~ jan$past_median)

Residuals:
   Min     1Q Median     3Q    Max 
-488.4  -25.5  -12.0    2.2 4220.9 

Coefficients:
                Estimate Std. Error t value Pr(>|t|)    
(Intercept)     19.66902   12.12200   1.623    0.105    
jan$past_median  0.91288    0.09119  10.011   <2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 181 on 735 degrees of freedom
Multiple R-squared:   0.12, Adjusted R-squared:  0.1188 
F-statistic: 100.2 on 1 and 735 DF,  p-value: < 2.2e-16

Taking out all gas rice less than 1 gwei.

Assuming these low transactions are miner transactions

Percentage of data that was taken out.

jan2 <- which(jan$gasprice < 1)
jan2 <- jan[-jan2,]
jan2 <- scale(jan2)
jan2 <- as.data.frame(jan2)
sum(jan$gasprice < 1)/length(jan$gasprice) #percentage of data we took out

[1] 0.001356852

janmodel2 <- lm(jan2$gasprice~jan2$past_median)

Summary of Model with out bids less than one gwei

summary(janmodel2)

Call:
lm(formula = jan2$gasprice ~ jan2$past_median)

Residuals:
    Min      1Q  Median      3Q     Max 
-2.5320 -0.1330 -0.0630  0.0104 21.8758 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      6.203e-17  3.460e-02    0.00        1    
jan2$past_median 3.465e-01  3.462e-02   10.01   <2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.9387 on 734 degrees of freedom
Multiple R-squared:  0.1201,    Adjusted R-squared:  0.1189 
F-statistic: 100.1 on 1 and 734 DF,  p-value: < 2.2e-16

Taking out all bids over 700 Percentage of data taken out

jan3 <- which(jan$gasprice > 700)
jan3 <- jan[-jan3, ]
jan3 <- scale(jan3)
jan3 <- as.data.frame(jan3)
sum(jan$gasprice > 700)/length(jan$gasprice)

[1] 0.006784261

janmodel3 <- lm(jan3$gasprice~jan3$past_median)

Summary of model without bids less than one gwei and without bids over 700

summary(janmodel3)

Call:
lm(formula = jan3$gasprice ~ jan3$past_median)

Residuals:
    Min      1Q  Median      3Q     Max 
-5.8906 -0.2174 -0.0322  0.1633  5.3426 

Coefficients:
                   Estimate Std. Error t value Pr(>|t|)    
(Intercept)      -1.987e-16  2.330e-02    0.00        1    
jan3$past_median  7.767e-01  2.331e-02   33.31   <2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.6303 on 730 degrees of freedom
Multiple R-squared:  0.6032,    Adjusted R-squared:  0.6027 
F-statistic:  1110 on 1 and 730 DF,  p-value: < 2.2e-16

Plot of the original data gas price vs. past median block

plot(janscale$gasprice, janscale$past_median)

Plot of gas price vs. past median block without fees under 1 and fees over 700

plot(jan3$gasprice, jan3$past_median)