Power BI自定义R可视化（PBIVIZ）动态绘图报错求助

Hey there! Let's work through your Power BI custom R visual (PBIVIZ) problems together—both fixing that "Values object does not exist" error and building the dynamic line chart you need.

1. Fixing the "Values object does not exist" Error #

The 3-year-old tutorial you're using is likely referencing an older version of the PBIVIZ R template. Power BI has updated how it passes data to custom R visuals:

• Older templates used a dedicated Values object to hold bound fields, but modern Power BI custom R visuals now use a default data frame called dataset for all your bound fields.
• Replace every instance of Values in your code with dataset. For example, if you had df <- Values, change it to df <- dataset.
• Double-check your pbiviz.json file to ensure your data roles are configured correctly—this ensures fields are properly loaded into the dataset frame.

2. Building a Dynamic Line Chart (No Hardcoded Column Names) #

To support free field switching without hardcoding column names, we'll rely on the order of bound fields and flexible data manipulation. Here are two solid implementations:

Option 1: Base R (Lightweight, No Extra Packages) #

# Exit early if no data is bound
if(nrow(dataset) == 0) return()

# Assume first bound field is X-axis (category/time), rest are Y-axis values
x_axis <- dataset[, 1]
y_series <- dataset[, -1]

# Initialize plot with first series
plot(x_axis, y_series[,1], type = "l", 
     col = 1, xlab = names(dataset)[1], ylab = "Value")

# Loop through remaining series to add lines
for(i in 2:ncol(y_series)){
  lines(x_axis, y_series[,i], col = i)
}

# Add dynamic legend using column names
legend("topright", legend = names(y_series), 
       col = 1:ncol(y_series), lty = 1)

Option 2: ggplot2 (Polished, Flexible Visuals) #

This uses tidyr to reshape data for easier multi-series plotting:

library(ggplot2)
library(tidyr)

# Exit early if insufficient data
if(nrow(dataset) == 0 || ncol(dataset) < 2) return()

# Reshape wide data to long format (works for any number of Y-series)
long_data <- pivot_longer(dataset, 
                          cols = -1,  # Exclude first column (X-axis)
                          names_to = "Series", 
                          values_to = "Value")

# Build dynamic plot
ggplot(long_data, aes(x = .data[[names(dataset)[1]]], y = Value, color = Series)) +
  geom_line(linewidth = 1) +
  labs(x = names(dataset)[1], y = "Value") +
  theme_minimal()

Critical Setup for Dynamic Behavior #

Update your pbiviz.json to define clear data roles that let users bind fields flexibly:

"dataRoles": [
  {
    "displayName": "Category (X-axis)",
    "name": "category",
    "kind": "Grouping"
  },
  {
    "displayName": "Values (Y-axis)",
    "name": "values",
    "kind": "Measure",
    "supportsMultiple": true
  }
]

• Users will bind their X-axis field to the "Category" role, and any number of numeric fields to "Values"—your code will automatically adapt to their selections.

3. Quick Troubleshooting Tips #

• Add print(str(dataset)) at the start of your R code to inspect the structure of the data being passed—this helps confirm fields are loading correctly in Power BI's debug window.
• Ensure all required R packages (like ggplot2 or tidyr) are installed in your Power BI R environment, and load them explicitly with library() in your code.

内容的提问来源于stack exchange，提问作者Fehnraal