EDIT 1
To clarify, why there is such a huge difference, the value of a subject can change and there were also new subjects in 2018 included which made some subjects of 2017 irrelevant. So I think there are 2 things to analyse:
- how a constant subject changed over time
- how a subject which got changed, changed over time
at the moment I am mixing these 2 up... therefor i don't get any good data...
EDIT 2:
I thought this problem through again. I think I have to many "unstable" attributes in this data set.
- Value of the items changes
- new items and removed items
- count of sellers
- count of customers
all this attributes change every year. So the real question i have to answer first is, how to show a representative diagram with this data set.
If i can answer this question, i won't have the issue i described below.
I try to show a statistic where it displays the deviation of 2 values in 2017 and 2018.
For this i use the formula:
$$ \frac {(x-y)}{x} *(-1)=?\\ $$
a simple exampel:
$$ \frac {(20-50)}{20} *(-100)= 150\% \\ $$
so i see, there was a 150% increase. Most of the data works just fine with this.
Nevertheless I'm unsure in two cases.
- When 2018 has increased a lot
- When 2018 has decreased a lot
1. $$ \frac {(20-18'000'000)}{20} *(-100) \approx 89'999'900\% \\ $$
2. $$ \frac {(18'000'000-20)}{18'000'000} *(-100)\approx -99.9\% \\ $$
As you can see, my Problem is, the formula is at the moment not informative. I have no clue how to adjust the axis so that i can show this data any useful. Did i made a mistake with my thinking process?