|
Hello, It appears that log scaling is not working properly for this chart. It works for others. Thank you.
Logarithmic scaling is a feature that changes where the numbers of the vertical scale of a chart fall. On "normal" charts (i.e. charts with "Arithmetic" scaling) each number on the vertical scale is an equal distance from the previous and next numbers. For example, an arithmetic scale chart that goes from 20 to 40 vertically would have 30 (the middle number) positioned exactly halfway up the vertical scale.
Logarithmic scaling locates the numbers of the vertical scale based on their percentage increase rather than their actual value. So, in the previous example, since 30 is 75% of the top value (40), 30 is located 3/4th of the way up the vertical scale on a log chart.
Log charts are great for looking at situations where the percentage gain or loss is more important that the actual price levels - very long-term charts for example.
Now, getting back to the question, the is one "problem" with Log charts however. The percentage change from 0 to any positive number is... infinity. Ugh. Infinity cannot be charted (believe me we've tried! ;-) ). Therefore, any data sets with a zero in them (or any datasets that cross the zero line) cannot be plotted on a log scale chart.
Whenever we get a request for a log scale chart of a dataset that contains or crosses zero, we automatically disable log scaling and using standard arithmetic scaling instead. That is what is happening in this case. Recent economy troubles have caused the number of new highs to dry up and zeros started appearing in $NYHGH. As soon as that happened, we were no longer able to plot it on a log scale chart.