Generated on your device · no upload, no signup
Log-log paper puts both axes on a logarithmic scale, so each square block of the grid represents a power-of-ten range on both dimensions rather than a fixed linear step. That's the layout to reach for when both variables in a relationship span multiple orders of magnitude — power-law relationships, allometric scaling in biology, particle size distributions, and many physics and engineering datasets all plot as straight lines on log-log paper where they'd be unreadable on a linear grid.
Set how many decades each axis spans — the tool generates a square grid of that many decades by that many decades, so a 3×3 setting covers three orders of magnitude in both X and Y. Choose US Letter or A4 and adjust line weight so the major decade boundaries stand out clearly from the internal log-scale subdivisions.
Plotting relationships where both variables span multiple orders of magnitude — power laws, scaling relationships, and datasets where a linear or semi-log grid would compress most of the data into an unreadably small area.
Match each axis to the range of that variable. If your X-values span 1 to 100 and Y-values span 1 to 10,000, you'd want roughly 2 decades on X and 4 on Y — though this version uses the same decade count for both axes, so pick whichever range needs the most room.
A straight line on log-log axes indicates a power-law relationship (y = ax^b) — the slope of the line gives you the exponent b, which is often the thing you're trying to measure from experimental data.
Yes — semi-log paper (one linear axis, one log axis) is available separately if only one of your variables is exponential.