In the past, there was a problem that when consecutive valid mountains <1> and <2> having peak values of the same polarity were detected for some reasons, the value of the shake count was reset even if the continuity between the valid mountains <1> and <2> have been confirmed.