Data modeling: R script error in Power Bi

[Qmars007]

Hello,

I try to run a simple regression script in Power BI but I get this message in Power BI:
DataSource.Error: ADO.NET: R script error. Error in eval(predvars, data, env) : object 'price' not found Calls: lm ... eval -> -> model.frame.default -> eval -> eval Execution halted Details: DataSourceKind=R DataSourcePath=R Message=R script error. Error in eval(predvars, data, env) : object 'price' not found Calls: lm ... eval -> -> model.frame.default -> eval -> eval Execution halted ErrorCode=-2147467259 ExceptionType=Microsoft.PowerBI.Scripting.R.Exceptions.RScriptRuntimeException

This simple regression script runs well in R Studio. What's wrong? Any idea?
Please find below the raw data. 

Many thanks

price	engine-size	horsepower
13495	130	111
16500	130	111
16500	152	154
13950	109	102
17450	136	115

[v-qiuyu-msft]

Hi @Qmars007,

You can modify your R script below: 

Regression <- lm(price ~ horsepower + enginesize, data=dataset)
#Show the results
summary(Regression)
with(dataset,plot(horsepower,enginesize))
abline(Regression)

Best Regards,
Qiuyun Yu 

[Qmars007]

Hi Qiuyun,

Thanks for your reply and suggestions. I was wondering with summary (regression) script, shouldn't I see the coefficients of the model in Power BI as below:

Residuals:
Min 1Q Median 3Q Max
-11180 -1908 -201 1629 13277

                     Coefficients:               Estimate Std. Error         t value              Pr(>|t|)
(Intercept)    -8510.47                    843.84                           -10.085            < 2e-16 ***
horsepower   60.21                        12.24                                4.917               1.85e-06 ***
enginesize     122.44                      11.01                              11.118             < 2e-16 ***
----
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 3678 on 196 degrees of freedom
Multiple R-squared: 0.7896, Adjusted R-squared: 0.7875
F-statistic: 367.8 on 2 and 196 DF, p-value: < 2.2e-16
-----------------------------------------------------------------
If that is not feasable in Power BI please let me know
Best

Qmars

[v-qiuyu-msft]

Hi @Qmars007,

The summary (regression) result you mention is returned in R studio Console pane not in Plots pane. You need to ensure the result can be plot in Plots pane in R studio, then you can use the R script in Power BI to visualize data. For example, in my previous post, I use abline() to plot a chart. 

Best Regards,
Qiuyun Yu 

[Qmars007]

Hi Qiuyuan,
Thanks for your prompt response.
My aim is to display the summary of my regression analysis in Power BI as it does appear in R studio pane rather than the graph of the correlation between the two explanatory variables.
How can this be done? 
Thanks very much
Qmars

[v-qiuyu-msft]

Hi @Qmars007,

Till now, it's almost the R issue instead of Power BI visual. As I mentioned before, you might need to write R script to ensure Summary(regression) result display in Plots pane within R studio. Then you can use the same script in Power BI desktop to plot the visual. 

Best Regards,
Qiuyun Yu 

[Qmars007]

Thanks for your message. Would you be able to let me know "how to write R script to ensure Summary(regression) result display in Plots pane within R studio" in a way that I can see the executive summary of regression analysis? Kind regards Qmars

[v-qiuyu-msft]

Hi @Qmars007,

Would you please share the R script? 

If possible, please share pbix file as well. You can upload it to your OneDrive and send the share link to me via private message. 

Best Regards,
Qiuyun Yu