Skip to content

Commit

Permalink
Fixing typos
Browse files Browse the repository at this point in the history
  • Loading branch information
kaz-Anova committed Sep 7, 2017
1 parent 2dca889 commit 9e90d8a
Show file tree
Hide file tree
Showing 2 changed files with 5 additions and 4 deletions.
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -477,7 +477,7 @@ Java -jar stacknet.jar **_predict_** **sparse**=false **has_head**=true **model*
- [Kaggle-TwoSigma](/example/twosigma_kaggle/EXAMPLE.MD)
- [Kaggle-TwoSigma Random Forest using the Library](/example/twosigma_kaggle_java_rf/EXAMPLE.MD)
- [Kaggle-Amazon Classification challenge and use of data_prefix](/example/example_amazon/EXAMPLE.MD)
- [Kaggle-Zillow regerssion example]((/example/zillow_regression_sparse/EXAMPLE.MD)
- [Kaggle-Zillow regerssion example](/example/zillow_regression_sparse/EXAMPLE.MD)



Expand Down
7 changes: 4 additions & 3 deletions example/zillow_regression_sparse/README.MD
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ To run follow the next steps:
2. Clone the git as `git clone https://github.com/kaz-Anova/StackNet.git`
3. Ensure you have **Python** installed and that Python can be found on **PATH** as StackNet now makes python calls as subprocesses. Also ensure that **sklearn 0.18** or above is installed. If you have trouble doing it in windows, see [this](https://www.youtube.com/watch?v=Y2q_b4ugPWk).
4. Make certain you have **Java** higher than 1.6 installed and that Java is in your **PATH**. Have a look at [this](https://www.java.com/en/download/help/path.xml) if you encounter trouble.
5. **Either** create a new (base) directory where you will put in the ** StackNet.jar** file AND the **lib\** folder **OR** just work inside the cloned github repo. Whichever one you chose, now create an `input` folder where you put all data from (1). Additionally copy and all files contained in this example (**not inside** the input folder, just in the base directory where you will execute the code from). In principle the only thing you need to run StackNet is the ** StackNet.jar** file and the **lib\** folder. The **lib/** folder contains the binaries to run some of the newer wrappers (xgboost, lightgbm, fastrgf).
5. **Either** create a new (base) directory where you will put in the ** StackNet.jar** file AND the **lib/** folder **OR** just work inside the cloned github repo. Whichever one you chose, now create an `input` folder where you put all data from (1). Additionally copy and all files contained in this example (**not inside** the input folder, just in the base directory where you will execute the code from). In principle the only thing you need to run StackNet is the ** StackNet.jar** file and the **lib/** folder. The **lib/** folder contains the binaries to run some of the newer wrappers (xgboost, lightgbm, fastrgf).
6. From the base directory execute the **make_stacknet_data.py** as `python make_stacknet_data.py` to generate 2 files in sparse format (*dataset2_train.txt,* *dataset2_test.txt*) . This is taken from [here](https://www.kaggle.com/danieleewww/xgboost-lightgbm-and-olsv107-w-month-features). It is the middle dataset in that code (the one that applies outlier removal).
6. Optionally, it will be good to do some sanity checks to ensure that all binaries work fine for [xgboost](https://github.com/kaz-Anova/StackNet#install-xgboost), [lightgbm](https://github.com/kaz-Anova/StackNet#install-lightgbm), [fast_rgf](https://github.com/kaz-Anova/StackNet#install-fast_rgf-new) and [python algos](https://github.com/kaz-Anova/StackNet#install-sklearn-algorithmsnew) . There is high chance linux users will need to `chmod +x` the executables inside the lib/[your_os]/[tools_name]/
7. **Run in the command line**. You may need to look at the [parameters’](https://github.com/kaz-Anova/StackNet#command-line-parameters) section on GitHub to understand more about the available models and their hyper paramaters. From the base directory run
Expand All @@ -20,7 +20,8 @@ To run follow the next steps:

The architecture of this ensemble is illustrated bellow.

![Alt text](/images/structure.png?raw=true "architecture")
![Alt text](/example/zillow_regression_sparse/images/structure.png?raw=true "architecture")


10. your predictions are now saved in **pred2.csv** in the base directory.
11. You can generate a submission via executing the **create_submission.py** as `python create_submission.py`. It will be saved to *output_dataset2.csv*
Expand All @@ -30,5 +31,5 @@ The architecture of this ensemble is illustrated bellow.

This is how your base directory needs to look like after everything is done:

![Alt text](/images/tree.png?raw=true "architecture")
![Alt text](/example/zillow_regression_sparse/images/tree.png?raw=true "architecture")

0 comments on commit 9e90d8a

Please sign in to comment.