alignment-prototype

Code base for a prototype for the hackathon on curriculum structure alignments.

Operations

These steps should work for any of the users with sudo rights on the server, including aron, ivan, jamie, kevin, kevino, and richard.

Deploy

ssh 35.235.65.36
    sudo su
        cd /projects/alignment-prototype/
        git pull
        systemctl restart gunicorn

Export data

ssh 35.235.65.36
    sudo su kevino
        cd /projects/alignment-prototype/
        pipenv run alignmentpro/manage.py exportdata

TODOs

How do we make robust/forgiving/helpful errors when bad Excel formatting?

Install

pipenv --python python3.6 install --dev

Services

Assuming you have Postgres DB running on localhost, you'll need to create the DB called alignmentpro using the following command:

# make sure postgres is running
createdb alignmentpro
createdb standards

Setup

Frontend

yarn install
yarn run build

Backend

./alignmentpro/manage.py makemigrations alignmentapp  # in case some not commited
./alignmentpro/manage.py migrate

Load data fixutres

./alignmentpro/manage.py loaddata alignmentpro/alignmentapp/fixtures/paremeters.json
./alignmentpro/manage.py loaddata alignmentpro/alignmentapp/fixtures/subject_areas.json
./alignmentpro/manage.py loaddata alignmentpro/alignmentapp/fixtures/admin_user.json

./alignmentpro/manage.py loaddata imports/curriculumdocuments/CCSSM.json
./alignmentpro/manage.py loaddata imports/curriculumdocuments/NGSS.json
./alignmentpro/manage.py loaddata imports/curriculumdocuments/Australia.json
./alignmentpro/manage.py loaddata imports/curriculumdocuments/khan_academy_us.json
./alignmentpro/manage.py loaddata imports/curriculumdocuments/CA-CTE.json

./alignmentpro/manage.py runserver

Load sample data

We've prepared some sample curriculum structures in this gsheet.

./alignmentpro/manage.py importchunk \
    --source_id='sample1-uganda-biology' \
    --title='Uganda biology sample topic' \
    --country='Uganda' \
    --digitization_method='scan_manual' \
    --gsheet_id='1-ei7BBMOx0udbXxyLJjMPYLW0EJWg9wFUyV9ODa8m5o' \
    --gid='1733263132'

./alignmentpro/manage.py importchunk \
    --source_id='sample2-kenya-math' \
    --title='Kenya sample math topics' \
    --country='Kenya' \
    --digitization_method='scan_manual' \
    --gsheet_id='1-ei7BBMOx0udbXxyLJjMPYLW0EJWg9wFUyV9ODa8m5o' \
    --gid='1031912854'

./alignmentpro/manage.py importchunk \
    --source_id='sample4-uganda-chemistry' \
    --title='Uganda chemistry sample topic' \
    --country='Uganda' \
    --digitization_method='scan_manual' \
    --gsheet_id='1-ei7BBMOx0udbXxyLJjMPYLW0EJWg9wFUyV9ODa8m5o' \
    --gid='1069644580'

Interactive debug

./alignmentpro/manage.py shell

Data export

Use the command

./alignmentpro/manage.py exportdata

The option --drafts will also export draft curriculum docs, while specifying --includetestdata will include the judgments test data.

Printing trees

Export tree by source_id

./alignmentpro/manage.py printtree --source_id=<source_id>

Use this command to export all Australia curriculum documents:

./alignmentpro/manage.py printtree  --short_identifiers --country=Australia

Importing chunkedscans folder hierarchy

Assuming the folder files/chunkedscans/KICDvolumeII_KICD secondary curriculum volume II contains the sample topic chunks, use the following command to import them into DB:

./alignmentpro/manage.py importsectionzips --sourcedir='files/chunkedscans/KICDvolumeII_KICD secondary curriculum volume II'

This will create a document with source id KICDvolumeII and title KICD secondary curriculum volume II, add a root DocumentSection with the same name as the document KICD secondary curriculum volume II, then recursively create DocumentSection for each of the subfolders. NOTE: this script only works for the filenames of the KICD curriculum.

Sample documents, nodes, and human judgment edge

from alignmentapp.models import CurriculumDocument, StandardNode
from alignmentapp.models import HumanRelevanceJudgment

d1, _ = CurriculumDocument.objects.get_or_create(title='KICD Math', source_id='kicd-math-sample')
n1 = StandardNode.add_root(title='KICD standards root', document=d1)
n2 = n1.add_child(title='Math', document=n1.document)
n3 = n2.add_child(title='Algebra', document=n2.document)
n4 = n2.add_child(title='Geometry', document=n2.document)

d2, _ = CurriculumDocument.objects.get_or_create(title='Uganda Math', source_id='uganda-math-sample')
n5 = StandardNode.add_root(title='NCDC Math standard root', document=d2)

e = HumanRelevanceJudgment(node1=n1, node2=n2, rating=0.7)
e.save()


print( n1.judgments.all() )   # judgments searches both node1 and node2 positions of n1
# <QuerySet [<HumanRelevanceJudgment: <StandardNode:  KICD Math standard root> <--0.7--> <StandardNode:  Uganda root node>>]>

print( n2.judgments.all() )
# <QuerySet [<HumanRelevanceJudgment: <StandardNode:  KICD Math standard root> <--0.7--> <StandardNode:  Uganda root node>>]>

Clean start

Delete all data and start form clean slate:

dropdb alignmentpro
rm -rf alignmentpro/alignmentapp/migrations/00*py
createdb alignmentpro
sleep 1
./alignmentpro/manage.py makemigrations alignmentapp
./alignmentpro/manage.py migrate

jamalex/alignment-prototype