An open API service indexing awesome lists of open source software.

https://github.com/coatless/stat429-fa15-autograder

Autograder results from STAT 429 Fall 2015
https://github.com/coatless/stat429-fa15-autograder

Last synced: 3 months ago
JSON representation

Autograder results from STAT 429 Fall 2015

Awesome Lists containing this project

README

          

Autograder Leaderboard for STAT 429 in Fall 2015
================================================

Problem 2
---------

| Position| Team ID | MAPE| Submission ID | Number of Submissions|
|---------:|:--------|-------:|:-----------------|----------------------:|
| 1| 6 | 0.0806| 1516acb4425ab176 | 43|
| 2| 12 | 0.0889| 1516b5888697a0bf | 45|
| 3| 1 | 0.1882| 15168a2904506adb | 52|
| 4| 9 | 0.2060| 1516b8db4123b453 | 42|
| 5| 2 | 0.2113| 15168888ffdd1f5d | 12|
| 6| 4 | 0.2130| 151654139af4f930 | 15|
| 7| 11 | 0.2197| 151545605ad01306 | 6|
| 8| 10 | 0.2276| 1516b3749e8ca9a3 | 32|
| 9| 8 | 0.2370| 15163b5d36202dc6 | 8|
| 10| 5 | 0.2424| 15163c20867d24d6 | 14|
| 11| 7 | 0.2496| 1516afdeb596373f | 28|
| 12| 3 | 0.2664| 151659e5c4d73f1d | 18|

Problem 3
---------

| Position| Team ID | MAPE| Submission ID | Number of Submissions|
|---------:|:--------|-------:|:-----------------|----------------------:|
| 1| 1 | 0.0073| 151651ebf5b99372 | 6|
| 2| 6 | 0.0074| 151647155d43e02b | 23|
| 3| 2 | 0.0075| 15168528c4c15ec2 | 13|
| 4| 11 | 0.0078| 151610c99ac345ca | 16|
| 5| 5 | 0.0086| 151601762e0604c0 | 12|
| 6| 7 | 0.0093| 15166160f7ba0ffc | 18|
| 7| 10 | 0.0094| 1516b4aaf0f2301d | 33|
| 8| 12 | 0.0097| 15165e832caaa1d3 | 30|
| 9| 4 | 0.0098| 1516a0b07b2d1054 | 34|
| 10| 3 | 0.0101| 1515bc76a3d77f75 | 8|
| 11| 9 | 0.0152| 1516a9d5da1af1fc | 12|
| 12| 8 | 0.0167| 1513781f7e0d6768 | 11|

Problem 4
---------

| Position| Team ID | MAPE| Submission ID | Number of Submissions|
|---------:|:--------|-------:|:-----------------|----------------------:|
| 1| 12 | 0.2723| 1516891e2b6873a5 | 82|
| 2| 9 | 0.2936| 1516b39496fce462 | 36|
| 3| 4 | 0.2962| 151643f9ab4b9131 | 9|
| 4| 6 | 0.2963| 15166bc3438a6ff2 | 56|
| 5| 7 | 0.3148| 1516968193eb5f2a | 36|
| 6| 1 | 0.3174| 15169e7b3a2c8915 | 12|
| 7| 8 | 0.3242| 1516ae1af867a9de | 12|
| 8| 11 | 0.3264| 15165a42c8e91df0 | 22|
| 9| 10 | 0.3451| 1516096a9d3bd4b0 | 23|
| 10| 5 | 0.3467| 1516112adde56fb1 | 13|
| 11| 2 | 0.3588| 151698d5c6aba8c0 | 15|
| 12| 3 | 0.3798| 1515aca51e743585 | 6|

Autograder Submission Statistics
================================

Below are submission statistics regarding the current state the Autograder is in.

Overall Submissions
-------------------

| Team ID | Total Submissions|
|:--------|------------------:|
| 1 | 70|
| 10 | 88|
| 11 | 44|
| 12 | 157|
| 2 | 40|
| 3 | 32|
| 4 | 58|
| 5 | 39|
| 6 | 123|
| 7 | 82|
| 8 | 31|
| 9 | 90|

Submissions by problem
----------------------

| Problem ID | Total Submissions|
|:-----------|------------------:|
| 2 | 315|
| 22 | 1|
| 3 | 216|
| 4 | 322|

Problem State Status
--------------------

| Assignment Code | Total|
|:----------------|------:|
| ERROR | 142|
| RAN | 609|
| STORE | 64|
| TIMEOUT | 39|

- **STORE**: If multiple submissions for a given problem occur within the collection window, the autograder evalutes the newest and sends the oldest to storage.
- **RUN**: The code is in the Autograder's grading queue.
- **ERROR**: The Autograder encountered an error when attempting to run the code.
- **RAN**: The code has been evaluated and the results have been sent.
- **TIMEOUT**: The runtime of the code has exceeded the allotted amount of processing time given to each submission.

Details of the Autograder
=========================

The results obtained above were generated by a script written to autograde R code written by James Joseph Balamuta (balamut2 [at] illinois [dot] edu). The script was deployed in Prof. Stephane Guerrier's STAT 429 Time Series Analysis course in Fall 2015 at the University of Illinois at Urbana-Champaign (UIUC). The hope is to increase the adoption of this script across all modeling or programming method courses within the Statistics department at UIUC.

The script takes in a team's code submission to create a model and then evaluates the model's effectiveness using a 1-step forward prediction aided by a hidden testing data set. The teams are then assessed based on their MAPE Score. The lower the MAPE score, the better the team's score.

[MAPE](https://en.wikipedia.org/wiki/Mean_absolute_percentage_error) is defined to be:

``` r
mean(abs((Y - Y_hat)/Y))
```

After the competition is over, a post will detail the intricacies of the system and provide guidance on how to replicate it.