Using Histogram one can check data spread, estimate type of distribution (e.g. normality, although normality should be checked with normality test) and symmetry. In my on-line Histogram tool one can do one histogram of all data or multiple histograms - one for each level. Enter the data in the "Data in txt format" window (text area), where a single row contains one, two or three data. The first row can be the data header. For example, you can copy 1-3 columns from Excel and paste into the data window.

If the user does not provide his own data, the program makes a diagram from the default data displayed in the "Data in txt format" window. Data can be copied e.g. from Excel, but also can be entered manually.

The default data are presented in the article History of some SPC. By default on the diagram there are also tolerance limits on the level: LSL = 65 and USL = 70, which can be switched on and off, see columns "Execute" and "Hide/Show".

Other recommendable data are available in CSV format. Just download them, open in Excel, copy them and paste them to the data window (about 15 seconds of work). Warning: note that the data separator in the window are spaces and tabs (which can cause problems in the header row or 2nd or 3rd column, when the string contains spaces).

The data structure must be as follows:

  
• 1st column - numerical data, e.g. number of defects per production lot, dimensions of parts, invoice processing time, etc.
• 2nd column (optional, called level) - text data, e.g. who performed the measurement, who processed the invoice, etc.
• 3rd column (optional, called level) - text data like in 2nd column.     Additional level relative to which we want to analyze something, e.g. machine number, time of the day, etc.
• 2nd and 3rd column are logically the same. Theoretically there can be many levels. Usually, 3 levels are used maximally.    I offer 2 levels, which is enough in 90% of cases.

Example included in downloadable data: John and Tom produce parts, every 10th part is measured. The measurement result is in the 1st column. The operator's name is in the 3rd column. In addition, we know that John and Tom work once in the Morning and once in the Evening. In the 3rd column we store the operators' time of the day, i.e. Morning or Evening. Having this data, we can make several combinations of calculations, just change the option in the "Levels in column(s)".

We can make the following combinations of calculations by selecting the below presented options:

  
• "-" - only data from the 1st column. One Histogram will be created from all data.
• "2" - analysis of the 1st column in relation to the 2nd column. We will receive 2 Histograms: one for John and the other for Tom. One can compare the work of operators.
• "3" - analysis of the 1st column in relation to the 3rd column. We will receive 2 Histograms: one for Morning and the other for Evening. One can compare the effect of working time on the results.
• For the options "2" and "3" Histogram are sorted according to the sequence of the level names in the data.
• "2.3" - analysis of the 1st column in relation to the 2nd and 3rd column. We will receive 4 Histograms: "John|Evening", "John|Morning", "Tom|Evening", "Tom|Morning".    In this case it is easier to compare how John works at different times and the same comparison for Tom.
• "3.2" - analysis of the 1st column in relation to the 3rd and 2nd column. We will receive 4 Histograms: "Evening|John", "Evening|Tom", "Morning|John", "Morning|Tom".    In this case it is easier to compare how John and Tom work at the same time.

Regardless of the language chosen, the diagram is formatted according to the English convention. This applies to the format of numbers and descriptions.

The numeric data must be in one of two formats:

• Decimal separator ",". Groups separators in numbers are unacceptable. The correct data is for example 1,234 or 34456,12. Invalid data 34 456,12 - blank space not allowed.
• Decimal separator ".". Groups separators in numbers are unacceptable. The correct data is e.g. 1.234 or 34456.12. Invalid data 34,456.12 - no comma as group separator or 34 456.12 - blank space not allowed.

The Histogram diagram consists of a several elements. Each element of the diagram can be hidden or shown (in the "Hide/Show" column) if it has been activated (selection in the "Execute" column). I encourage you to use this "Hide/Show" function, because it will allow you to better understand the diagram construction.

If you do not want to analyze data by levels (create just one Histogram), you can show also mean line. However if you decide to analyze by levels, the Legend and Figure caption will appear below diagram. In addition each separate Histogram Finally, what can be hidden or shown using "[-]H5", where H5 is number of selected Histogram.

User can set the number intervals in the field "No. of intervals". If the field is empty or "0", then the number of intervals i is calculated from the formula (1), where IQR is Q3-Q1 (details are in description of Box Plot on-line tool) and n is sample size.

At the moment in case of transferring Histogram to another program, I can only offer a screenshot. However, the diagrams are made in SVG vector format.