31-github-report.rmd

---
title: "Test Summary"
output: github_document
params:
  input_dir: './build'
  output_name: ''
---

```{r setup, echo=FALSE, include=FALSE, results='hide'}
# Config knitr
knitr::opts_chunk$set(echo = TRUE, comment = NA, prompt = FALSE)
# Load all build artifacts 
data_files <- list.files(params$input_dir, pattern = "\\.dat$", full.names = TRUE)
lapply(data_files, load, .GlobalEnv)
```

```{r report_name, echo=FALSE}
paste("Report Name:",params$output_name)
```

## Summary Statistics from Raw Data
* List of expected columns in the jmeter output
* List of labels selected for analysis (The kind of events to be analyzed)
* Summary statistics of elapsed time (ms) for each label.

```{r raw_data_stats, echo=FALSE}
cat("Input column names:\n", paste(expected_fields, sep=","))
cat("Value of 'label' used for analysis:\n", paste(target_labels, sep=","))
summaries
```

## Results of Correlation and Linear Modeling 
How strong is the linear correlation between time elapsed and the number of succesful events? This correlation is performed assuming the elapsed time for each event should be increasing as the cumulative number of events increases.
A correlation coefficient near zero (< 0.25) indicates no or very weak correlation.
```{r correlations}
cor_results
```

Answering the question, "Is the time elapsed significantly dependent on the number of successful events of the type being analyzed?"  For example, is the time elapsed dependent on the number of successful "PUT Perf Container" events?
```{r linear_mods}
lmodels
```

## Plots
Plotting the time elapsed (ms) for each event by the cumulative number of events.

```{r bin_plots, echo=FALSE}
for(i in 1:length(bin_plots)){print(bin_plots[[i]])}
```

```{r dot_plots, echo=FALSE}
for(i in 1:length(dot_plots)){print(dot_plots[[i]])}
```

###### Template
_This report was generated from the template, 31-github-report.rmd_