The evaluation metric for this competition is
[Root Mean Squared Logarithmic Error](https://www.kaggle.com/marknagelberg/rmsle-function)
. Because this value is not only an absolute value, this metric is one of the most resistant metric to outliers that's possible using simple methods. Thus it is the most fair measure to calculate the number of bikes on the given data.
The calculation of this measure is made as follows:
\[Root Mean Squared Logarithmic Error=\sqrt{ {1/n}\sum( { \log(\hat{y_0} + 1)−log(y_0 +1) )}^2 }\]
In Python you can calculate it as :
import math

#A function to calculate Root Mean Squared Logarithmic Error (RMSLE)
def rmsle(y, y_pred):
	assert len(y) == len(y_pred)
	terms_to_sum = [(math.log(y_pred[i] + 1) - math.log(y[i] + 1)) ** 2.0 for i,pred in enumerate(y_pred)]
	return (sum(terms_to_sum) * (1.0/len(y))) ** 0.5

rmsle(y, y_pred)
In R you can calculate it as :
rmsle  < - function(y,ypred) {
    return(sqrt(mean((log(y+1,2) - log(y_pred+1,2))^2))
}
Submission Format
For every row in the dataset, submission files should contain two columns: id and cnt. It should be a comma-delimited list.
The file should contain a header and have the following format:
id, cnt
1, 0
2, 0
3, 0
4, 0
etc.