How to Solve R Error in eval(predvars, data, env): object not found

If you try to use the predict() function and the column names in the test data frame do not match those in the data frame used to fit the model, you will raise the error in eval(predvars, data, env): object ‘x’ not found. The ‘x’ will be the column name that does not exist.

You can solve this error by checking the column names of the data frames using names, for example,

names(data)

Then you can rename the columns so that they match, for example:

names(data)[names(data) == 'column_name_to_change'] <- 'new_column_name'

This tutorial will go through the error in detail and how to solve it with code examples.

---

Example

Let’s look at an example of fitting a linear regression model to some data. First, we will look at the data frame, which contains height and weight measurements for ten people.

data <- data.frame(height=c(176, 200, 134, 150, 160, 180, 140, 190, 145, 155),
weight=c(80, 104.7, 47, 55, 62.4, 70, 85, 66, 120, 60))

data

 height weight
1     176   80.0
2     200  104.7
3     134   47.0
4     150   55.0
5     160   62.4
6     180   70.0
7     140   85.0
8     190   66.0
9     145  120.0
10    155   60.0

In our example, height is the predictor variable, and weight is the response variable.

Next, we will fit a linear regression model to the data and get the summary view of the model:

model <- lm(weight ~ height, data=data)
summary(model)

Call:
lm(formula = weight ~ height, data = data)

Residuals:
   Min     1Q Median     3Q    Max 
-21.39 -14.68 -10.41  11.94  49.10 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)
(Intercept)  37.7907    57.9825   0.652    0.533
height        0.2283     0.3528   0.647    0.536

Residual standard error: 23.64 on 8 degrees of freedom
Multiple R-squared:  0.04977,	Adjusted R-squared:  -0.06901 
F-statistic: 0.419 on 1 and 8 DF,  p-value: 0.5356

Now that we have a model, we can use it to predict the weights of five test subjects given their heights.

test_data <- data.frame(heights=c(143, 210, 120, 188, 158))

predict(model, newdata = test_data)

Let’s run the code to see what happens:

Error in eval(predvars, data, env) : object 'height' not found

The error occurs because the names for the predictor variable do not match. We can obtain the column names of each data frame using the names() function:

names(data)
names(test_data)

[1] "height" "weight"

[1] "heights"

We can see that test_data has the column name heights, and data has the column name height.

Solution

We can solve this error by renaming the test_data column as follows:

names(test_data)[names(test_data) == 'heights'] <- 'height'
test_data

  height
1    143
2    210
3    120
4    188
5    158

Now that we have the correct column name, we can use the predict() function to predict the response values given the predictor values:

predict(model, newdata = test_data)

      1        2        3        4        5 
70.44321 85.74195 65.19141 80.71848 73.86830 

We successfully predicted the response values.

Summary

Congratulations on reading to the end of this tutorial!

For further reading on R related errors, go to the articles: 

• How to Solve R Error in lm.fit: na/nan/inf
• How to Solve R Error: Subscript out of bounds
• How to Solve R Error in apply: dim(X) must have a positive length
• How to Solve R Error in eval(predvars, data, env): object not found

Go to the online courses page on R to learn more about coding in R for data science and machine learning.

Have fun and happy researching!