Skip to content

Latest commit

 

History

History
79 lines (53 loc) · 2.28 KB

retrieving_hw.md

File metadata and controls

79 lines (53 loc) · 2.28 KB
Raw

Obtaining the Homework

If you are using the university provided Jupyter instance (here) rather than your own personal computer, the homeworks are stored in the folder class_materials. This folder will update every time you log in with any new homeworks or changes.

Working on the files on the JupyterHub

You can load the homework by clicking on class_materials/homeworks/homework-NeuralNet (or whatever the name is) in the top directory of your JupyterHub instance.

IMPORTANT: When you use notebooks on the NYU Jupyterhub you need to change the default kernel of the notebook. First open a notebook. Then selected "Kernel" from the menu bar. Then select "Change Kernel" and then select "Python [conda env:ccm]" from the list. Otherwise your python environment may not have the necessary packages to do the homework. In any case please always verify that the current notebook you are running says "Python [conda env:ccm]" in the upper right corner of the interface.

If you are running your own Jupyter instance use the following instructions to download the homeworks.

Retrieving Homework from Github to Jupyter

Open a terminal in Jupyter (Go to "new" in the upper right, then choose "terminal").

Cloning the repository

For first retrieval, clone the repository using

git clone https://github.com/brendenlake/CCM-site.git

Updating the repository

To update the repository, enter the repository folder using

cd CCM-site

and pull new commits using

git pull

Creating a copy of a HW

To create a copy of a HW problem set to work on in the home folder (and prevent possible merge conflicts on future pulls), run

cd ~
cp -r PATH/TO/HW .

For example, for homework 1 you would type

cp -r CCM-site/homeworks/homework-NeuralNet .

(don't forget trailing .)

Packages

The following python packages are installed on the NYU Jupyterhub and so if you want to create a similar setup please install the following using your preferred pacakge manager (anaconda, pip, etc..). The latest versions should be fine:

  • pandas
  • wikipedia
  • colorlover
  • jupyter_contrib_nbextensions
  • matplotlib
  • pingouin
  • scikit-learn
  • scikit-image
  • ptitprince
  • pomegranate
  • hmmlearn
  • nilearn
  • gym
  • torch
  • torchvision