# Histogram

File:Histogram example.svg
Example of a histogram of 100 normally distributed random values.

In statistics, a histogram is a graphical display of tabulated frequencies. A histogram is the graphical version of a table that shows what proportion of cases fall into each of several or many specified categories. The histogram differs from a bar chart in that it is the area of the bar that denotes the value, not the height, a crucial distinction when the categories are not of uniform width (Lancaster, 1974). The categories are usually specified as non-overlapping intervals of some variable. The categories (bars) must be adjacent.

The word histogram is derived from Greek: histos 'anything set upright' (as the masts of a ship, the bar of a loom, or the vertical bars of a histogram); gramma 'drawing, record, writing'. The histogram is one of the seven basic tools of quality control, which also include the Pareto chart, check sheet, control chart, cause-and-effect diagram, flowchart, and scatter diagram. A generalization of the histogram is kernel smoothing techniques. This will construct a very smooth Probability density function from the supplied data.

## Examples

As an example we consider data collected by the U.S. Census Bureau on time to travel to work (2000 census, [1], Table 5). The census found that there were 124 million people who work outside of their homes. People were asked how long it takes them to get to work, and their responses were divided into categories: less than 5 minutes, more than 5 minutes and less than 10, more than 10 minutes and less than 15, and so on. The tables shows the numbers of people per category in thousands, so that 4,180 means 4,180,000.

The data in the following tables are displayed graphically by histograms. An interesting feature of both diagrams is the spike in the 30 to 35 minutes category. It seems likely that this is an artifact: half an hour is a common unit of informal time measurement, so people whose travel times were perhaps a little less than, or a little greater than 30 minutes might be inclined to answer "30 minutes". This rounding is a common phenomenon when collecting data from people.

File:Travel time histogram total n.png
Histogram of travel time, US 2000 census. Area under the curve equals the total number of cases. This diagram uses Q/width from the table.
Data by absolute numbers
Interval Width Quantity Quantity/width
0 5 4180 836
5 5 13687 2737
10 5 18618 3723
15 5 19634 3926
20 5 17981 3596
25 5 7190 1438
30 5 16369 3273
35 5 3212 642
40 5 4122 824
45 15 9200 613
60 30 6461 215
90 60 3435 57

This histogram shows the number of cases per unit interval so that the height of each bar is equal to the proportion of total people in the survey who fall into that category. The area under the curve represents the total number of cases (124 million). This type of histogram shows absolute numbers.

File:Travel time histogram total 1.png
Histogram of travel time, US 2000 census. Area under the curve equals 1. This diagram uses Q/total/width from the table.
Data by proportion
Marks Scored No. of Students
10–20 1
20–30 1
30–40 4
40–50 4
50–60 8
60–70 7
70–80 11
80–90 6
90–100 5

This histogram differs from the first only in the vertical scale. The height of each bar is the decimal percentage of the total that each category represents, and the total area of all the bars is equal to 1, the decimal equivalent of 100%. The curve displayed is a simple density estimate. This version shows proportions, and is also known as an unit area histogram.

## Activities and demonstrations

The SOCR resource pages contain a number of hands-on interactive activities demonstrating the concept of a Histogram, histogram construction and manipulation using Java applets and charts.

## Mathematical definition

In a more general mathematical sense, a histogram is simply a mapping ${\displaystyle m_{i}}$ that counts the number of observations that fall into various disjoint categories (known as bins), whereas the graph of a histogram is merely one way to represent a histogram. Thus, if we let ${\displaystyle n}$ be the total number of observations and ${\displaystyle k}$ be the total number of bins, the histogram ${\displaystyle m_{i}}$ meets the following conditions:

${\displaystyle n=\sum _{i=1}^{k}{m_{i}}.}$

### Cumulative Histogram

A cumulative histogram is a mapping that counts the cumulative number of observations in all of the bins up to the specified bin. That is, the cumulative histogram ${\displaystyle M_{i}}$ of a histogram ${\displaystyle m_{i}}$ is defined as:

${\displaystyle M_{i}=\sum _{j=1}^{i}{m_{j}}}$

### Number of bins and width

There is no "best" number of bins, and different bin sizes can reveal different features of the data. Some theoreticians have attempted to determine an optimal number of bins, but these methods generally make strong assumptions about the shape of the distribution. You should always experiment with bin widths before choosing one (or more) that illustrate the salient features in your data.

The number of bins ${\displaystyle k}$ can be calculated directly, or from a suggested bin width ${\displaystyle h}$:

${\displaystyle k=\left\lceil {\frac {\max x-\min x}{n}}\right\rceil }$

The braces indicate the ceiling function.

Sturges' formula
${\displaystyle k=\lceil \log _{2}n+1\rceil }$

which implicitly bases the bin sizes on the range of the data, and can perform poorly if ${\displaystyle n<30}$.

Scott's choice[1]
${\displaystyle h={\frac {3.5s}{n^{1/3}}}}$

where ${\displaystyle h}$ is the common bin width, and ${\displaystyle s}$ is the sample standard deviation.

Freedman-Diaconis' choice[2]
${\displaystyle h=2{\frac {\operatorname {IQR} (x)}{n^{1/3}}}}$

which is based on the interquartile range

## Continuous data

The idea of a histogram can be generalized to continuous data. Let ${\displaystyle f\in L^{1}(R)}$ (see Lebesgue space), then the cumulative histogram operator ${\displaystyle H}$ can be defined by:

${\displaystyle H(f)(y)=\mu \{x:f(x)\leq y\}}$.

${\displaystyle \mu }$ is the Lebesgue measure of sets. ${\displaystyle H(f)}$ in turn is a real function. The (non-cumulative) histogram is defined as its derivative:

${\displaystyle h(f)=H(f)'}$.

For differentiable functions ${\displaystyle f}$ with only finitely many intervals of monotony this can be rewritten as

${\displaystyle h(f)(y)=\sum _{\xi \in \{x:f(x)=y\}}{\frac {1}{|f'(\xi )|}}}$.

${\displaystyle h(f)(y)}$ is undefined if ${\displaystyle y}$ is the value of a stationary point.

2. Freedman, David (1981). "On the histogram as a density estimator: ${\displaystyle L_{2}}$ theory". Zeitschrift für Wahrscheinlichkeitstheorie und verwandte Gebiete. 57 (4): 453–476. doi:10.1007/BF01025868. Unknown parameter |coauthors= ignored (help)