The data below shows the GDP of each country by year. Each year from 1960 to 2024 is represented as a column (variable). Therefore, each row represents the change in GDP of one country.
When you look at the data, you can see that there are missing values in places .
Data source: World Bank ( https://www.worldbank.org/ ) Date of acquisition: 2025/12/16
Possible reasons for missing values in this data include problems with the statistical collection system, or surveys and tabulations not yet being carried out. If you want to understand GDP trends by country or compare countries later, it is important to first understand the status of missing values in the data.
To check for missing values in your data, JMP provides a platform called "Check for Missing Values" that allows you to check for patterns in missing values and perform imputation.
But first,
- How many missing values are there?
- What pattern does it occur in?
If you want to quickly check the following, you can use other functions.
This article presents three ways to quickly and effectively check for missing data.
We recommend checking in the following order: 1 → 2 → 3.
- Use Column Manager: Find the number of missing values for each variable
- Use of "missing value pattern display": Understand trends in missing data
- Use of "cell plot": Understand missing data patterns relative to data order
Let's take a look at each of 1 to 3.
1. Using the Column Manager
In JMP 18 and later, you can instantly see summary statistics for each column by clicking the Column Manager button in the Columns panel (center left of the data table), which allows you to see the number of missing values in each column.
The data show that there are missing values in all years, with a particularly high number of missing values from around 1960 to 1964, and a gradual decrease thereafter.
2. Using the Missing Value Pattern Display
In 1, we checked the number of missing values for each variable, but there may be patterns of missing data, such as "data not being collected for two consecutive years."
The "Show Missing Value Patterns" feature allows you to check whether there is a pattern to the missing values in a data table. It can also be used to extract rows that do not contain any missing values.
If you run Table > Show Missing Value Patterns on the columns from 1960 to 2024 (65 columns in total), a new data table will be output. Here, we show the results sorted by frequency in descending order.
For values in the output table, 0 indicates not missing and 1 indicates missing.
The column "Pattern" indicates whether or not there are missing values for the specified column using 0/1. The column "Frequency" indicates the number of rows that correspond to that missing value pattern.
The first row has the pattern "000000…000", indicating that all columns have no missing values, meaning that values are entered for all years. There are 110 such rows in this data.
The fifth row shows a pattern where the first five columns (1960-1964) are missing data, but there are no missing data from 1965 onwards, and we can see that there are six such rows.
The table output by "Show Missing Value Patterns" is linked to the original data. For example, if you select the first row, the corresponding row will be selected in the original data table.
In this state, you can use Table > Subset to extract only data that does not contain any missing values.
3. Using "Cell Plot"
It is also important to understand where missing values occur: near the beginning, middle, or end of the data. For example, in process data acquired in chronological order, you may want to understand whether there are more missing values in the first half or the second half of the data.
In such cases, using "cell plot" allows you to intuitively visualize the distribution of missing values in a reduced cell display of the data.
In this example, we sort countries in descending order using the average GDP for the five years from 2020 to 2024, and examine the relationship between recent GDP levels and missing data.
Click Graph > Cell Plot and specify Year as the Y, Response Variable and Country Name as the Label. The Cell Plot report will then be displayed.
A cell plot is a graph that visualizes each cell in a data table as a colored rectangle.
Notice that instead of the cell color, you see an "X" in the cell , which indicates that the cell contains a missing value.
Looking at the distribution of "x"s, we can see that countries with lower GDP (those at the bottom of the data) tend to have more missing data, and that there are several countries with missing data for the recent years of 2023 and 2024.
One reason for this is thought to be that countries with higher GDPs have better systems in place to stably collect and publish such economic statistics.
Missing values are an important point to check before beginning any data analysis.
By understanding the presence or absence of missing data and trends in advance, subsequent analysis and interpretation can be carried out appropriately.
by Naohiro Masukawa (JMP Japan)
Naohiro Masukawa - JMP User Community
