Yes—but Excel’s built-in Histogram chart cannot put its horizontal category axis on a logarithmic scale. The practical workaround is to transform positive data with LOG10 and histogram those transformed values, or create custom logarithmic bins when you need bins labeled in the original units.
What a logarithmic histogram shows
A logarithmic scale is useful when values span several orders of magnitude—for example, measurements from 0.1 to 10,000—or when proportional differences matter more than equal numerical differences. On a base-10 scale, each step of one unit represents a tenfold change: 1, 10, 100, and 1,000 are evenly spaced in log space.
There are two related charts people may mean by “a logarithmic histogram.” One plots the logarithms of the observations, so equal-width bins represent equal ratios in the original data. The other groups the original observations into log-spaced intervals, such as 1–10, 10–100, and 100–1,000. These approaches are related, but their axis labels and interpretation differ.
Neither approach is appropriate for data containing zero or negative values without an explicit, justified transformation. The logarithm of zero or a negative number is undefined in the real-valued calculations Excel uses. Do not quietly replace such values with an arbitrary small positive number: that can change the apparent distribution.
Why Excel’s Histogram chart does not have a log x-axis
In Excel for Microsoft 365 on Windows, select the data and choose Insert > Insert Statistic Chart > Histogram to make a built-in histogram. Its horizontal axis is organized into bins, and its vertical axis shows frequencies. Excel provides histogram-specific options such as bin width, number of bins, and underflow or overflow bins, but the horizontal axis is a category axis rather than a numeric value axis with a logarithmic-scale setting.
Excel can offer a logarithmic scale for a chart’s value axis. That does not make the histogram’s horizontal bins logarithmic. A logarithmic frequency axis is also a different choice: it changes how bar heights are displayed, not how the observations are grouped. It cannot display zero on a log scale, which is a problem whenever a bin has no observations.
In short, changing an axis setting is not enough to turn the built-in histogram into a histogram with logarithmically spaced x-axis values. Choose the chart and binning method based on what you want the bars to mean.
Quick method: histogram the log-transformed data
This is usually the clearest option when the data are positive and you want to compare multiplicative ranges. With the original observations in A2:A1001, enter this formula in B2 and fill it down alongside the data:
=LOG10(A2)
For example, if A2 contains 100, the result is 2, because 10 to the power of 2 is 100. If the original values are 1, 10, and 100, their transformed values are 0, 1, and 2. A bin width of 1 in the transformed data therefore corresponds to a tenfold interval in the original data.
- Check the original column for zero, negative, blank, or text entries. Decide how to handle invalid or missing observations before charting.
- Calculate LOG10 in a new column, then select the resulting values.
- Choose Insert > Insert Statistic Chart > Histogram.
- Right-click the horizontal axis and choose Format Axis. Under Axis Options, set Bin width to 1 for decade-sized bins, or choose another width that suits the analysis.
- Label the chart and axis clearly, for example “log10(value)” and “Count.” Explain that one unit on the horizontal axis is a factor of ten in the original measurements.
The chart’s labels will be logarithms, not the original values. If a bin spans log10 values 2 to 3, its observations range from 100 up to, but not including, 1,000 when the bin edges are exact. That is mathematically valid, but readers may find original-unit labels easier to interpret.
Choosing bin widths and boundaries
Bin width determines what patterns the chart reveals. With log10-transformed data, a width of 1 groups values by powers of ten. A width of 0.5 groups them into factors of about 3.16, because 100.5 is approximately 3.16. A narrower width shows more detail but can make random variation look like structure; a wider width can conceal meaningful clusters.
Set boundaries deliberately when the values have meaningful thresholds. For instance, if measurements range from 0.1 to 10,000, decade boundaries at 0.1, 1, 10, 100, 1,000, and 10,000 are easy to explain. If you use a built-in histogram on transformed values, check the axis bounds and bin width so the bins actually align with the intended transformed boundaries.
- Use equal widths in log space when comparing multiplicative ranges or orders of magnitude.
- Use narrower log-space bins when the sample is large enough to support more detail.
- Use fewer, wider bins when observations are sparse or the chart is intended as a high-level summary.
- State the base and units in the chart title or axis label, especially if another reader might assume natural logarithms rather than base 10.
Custom bins labeled in the original units
If readers need to see ranges such as “10–100” rather than “1–2” on a log10 axis, make a frequency table using original-unit boundaries. For example, list lower boundaries in D2:D5 as 1, 10, 100, and 1,000, and upper boundaries in E2:E5 as 10, 100, 1,000, and 10,000. With source data in A2:A1001 and the first interval’s boundaries in D2 and E2, enter this in F2 and copy it down:
=COUNTIFS($A$2:$A$1001,">="&D2,$A$2:$A$1001,"<"&E2)
This counts values at or above the lower boundary and below the upper boundary. For instance, the 10–100 row includes 10 but excludes 100, which belongs to the next interval. For the final interval, decide whether to include the uppermost boundary; if necessary, use a separate final-bin calculation with a “less than or equal to” condition. Consistent boundaries prevent values on an edge from being counted twice or omitted.
Select the interval labels and counts, then insert a column chart using Insert > Insert Column or Bar Chart > 2-D Column. This produces equal-width categories labeled with the original ranges. It is a useful frequency display, but it is not a conventional histogram on a numeric logarithmic axis: each bar looks equally wide even though the original intervals have different numerical widths.
Counts, density, and what bar height means
When bin widths vary in the original units, a bar chart of raw counts can be misleading if it is presented as a conventional histogram. In a histogram, the bar’s area—not simply its height—represents the number or proportion of observations. A 10–100 bin is 90 units wide, while a 100–1,000 bin is 900 units wide. Giving both bars equal visual width and using raw counts does not show those unequal widths.
For a histogram on the original numerical scale with unequal-width bins, use frequency density as bar height: count divided by bin width. The bar area then equals the count. This can be difficult to build cleanly with Excel’s standard column chart because category columns have equal display widths. A chart of log-transformed observations avoids that particular mismatch by using equal-width bins in log space, but its density is per log unit, not per original unit. State which quantity the chart represents.
If the purpose is simply to compare how many observations fall in each named range, a frequency table and clearly labeled column chart may be entirely adequate. Avoid calling it a density histogram unless the interval widths and bar areas support that interpretation.
Formatting and interpreting the chart
Use a title that describes the data and transformation, such as “Distribution of log10 response values.” Include units for the original measurements somewhere on the chart or in a note. If the audience expects original-unit cutoffs, provide a small table mapping transformed boundaries back to original values—for example, log10 values 0, 1, 2, and 3 correspond to 1, 10, 100, and 1,000.
Excel’s built-in histogram can select bins automatically, but automatic choices are not necessarily meaningful for logarithmic analysis. After creating the chart, inspect the axis options and confirm that the range and width reflect your intended boundaries. Also check whether underflow or overflow bins are combining observations outside the displayed range. Those bins can conceal extreme values if their meaning is not made clear.
Do not interpret a taller bar as proof that values are more common per original measurement unit when the chart uses transformed data. It means more observations fell within that log-space interval. A decade interval from 10 to 100 covers 90 original units, while 100 to 1,000 covers 900; both have width 1 after a base-10 transformation.
Common problems and fixes
- #NUM! results: LOG10 cannot calculate a real result for zero or negative input. Find those records and handle them according to the data’s meaning rather than masking the error.
- Unexpected bin edges: Excel may choose automatic boundaries that do not line up with powers of ten. Set the bin width and bounds explicitly, or use a custom frequency table.
- Unclear axis labels: A horizontal axis labeled only with values such as 1 and 2 may be mistaken for the original scale. Label it as log10(value) and explain the corresponding original ranges.
- Empty bins on a log frequency axis: Zero counts cannot be represented on a logarithmic vertical axis. Keep the count axis linear or use a different display; adding an arbitrary count is not a neutral fix.
- Bars that imply the wrong widths: A column chart of custom categories gives each category the same visual width. Describe it as a count-by-range chart, or use a method that represents variable-width intervals appropriately.
Alternatives when you need a true log x-axis
For a quick, understandable distribution, histogram the LOG10 values and label the transformation. For counts in business-friendly original ranges, use a custom frequency table and clearly labeled column chart. If the analytical requirement is specifically a numeric x-axis with logarithmic spacing and bars whose widths represent the original intervals, Excel’s standard Histogram chart is not a direct solution. Consider specialist statistical software or a carefully constructed custom chart, and document whether bar height is count or density.
A scatter chart can use a logarithmic value axis for positive x-values, but it does not automatically become a histogram. It plots points rather than frequency bins. It may help display binned counts at selected positions, but that representation needs deliberate design and should not be confused with Excel’s built-in histogram.
Frequently Asked Questions
Can I make the x-axis of an Excel Histogram chart logarithmic?
No. The built-in Histogram chart uses a binned category axis and does not provide a logarithmic horizontal-axis setting. Transform the data first or create custom bins instead.
How do I make decade bins in Excel?
Apply LOG10 to positive observations and use a bin width of 1. The resulting intervals correspond to successive powers of ten, such as 10–100 and 100–1,000.
Can I include zero in a logarithmic histogram?
Not by applying LOG10 directly, because the logarithm of zero is undefined. Keep zero values separate or choose a transformation and analysis plan that is appropriate for the data; do not substitute a convenient number without justification.
Is a log-scale histogram the same as plotting counts on a log axis?
No. Transforming the observations changes the bins along the data dimension. A logarithmic frequency axis changes the displayed bar heights and cannot show zero counts.
Can Excel show original values after I transform them?
The built-in histogram labels the transformed values. You can add a mapping table or make a custom frequency chart with original-unit interval labels, while explaining that the latter uses equal-width categories visually.
Related
- Excel Hide Column Numbers? Hide Row and Column Headings
- Excel Hide Columns Based on Date: Methods That Work
- Excel Hide Columns From Certain Users: What Actually Works
Try it: the Percentage Formula Generator: percent of, percent change, increase and decrease, with the formula.
More in Charts & pivot tables
- How to Make a 4-Quadrant Graph in ExcelTo make a 4-quadrant graph in Excel, plot your paired X and Y values in an XY scatter…
- How to Insert a Treemap Chart in Excel: Step by StepTo insert a Treemap chart in Excel, arrange your data in columns from broad category to specific item,…
- Excel Chart Labels: Add, Format, and Fix ThemIn Excel for Microsoft 365 on Windows, select a chart, then choose Chart Design > Add Chart Element…
- Excel Combo Chart: How to Combine Bars and a Line in ExcelTo make a bar-and-line combo chart in Excel for Microsoft 365 on Windows, select your data, choose Insert…
- Excel Pivot Table Value Filter Not Working? FixesIf an Excel PivotTable value filter is not working, first confirm you applied it to a row or…
- Excel Graph Axis Range: How to Set Minimum and MaximumTo change an Excel graph’s axis range, select the chart axis, right-click it, choose Format Axis, and enter…
Sources: Function behaviour and menu paths: Microsoft Support documentation for Excel. Method & sources
Spotted an error in this guide? Let us know.