kekeke29341 / deepKarte

###Generating_karte This repository is for the purpose of research in medical records processing. ALL Data is not real data. They are generated by random from random dictionary.

###Elastic Search Use
Generate medical records for evaluating patient similarity
python make_dataset/es_sample_generate_validation_data.py NumberPatient INDEX_Name

###Flask $ Learning Usage

1. python flask/generate_textbox_html.py Speech/karte/karte_ep7.json
2. python flask/add_escape.py
3. python flask/app.py

###Text summarization
####Data Conversion (This part is important when you use Japanese data)

1. python src/strip_n.py text_summarization/Raw_data/Formatted_Training_Data.tsv
   (Sentences into each word)
2. python text_summarization/generate_data.py --input_dir=text_summarization/train --data_path=text_summarization/binary_train/NE109
3. python text_summarization/check_data.py --data_path=text_summarization/binary_train/NE109 --crc=4
   (Raw data -> train -> Binary_train) (If you want to convert google drive word file, first you convert it to Excel and A&B to TSV and then start from 1.)
   ####Generate Vocab dictionary
   python src/mecab_word_count.py text_summarization/data/national_exam109.txt

###Learning Start

1. python text_summarization/convert_data.py Speech/data/ep6/scripts.txt flask/learn/32.txt
2. python text_summarization/generate_data.py --input_dir=text_summarization/train --data_path=text_summarization/binary_train/output
   (Second argument is directory)

###Recording and Learning(+ Searching Realtime using Google Chrome)
sh shellscript/record.sh
python shellscript/main.py
python flask/app.py

###Recording and export Ruby on Rails(extraction method is completely diffrenet so src is in RB dir):
sh shellscript/record_voice_model2.sh
cd rails directory
rails s
rails runner app/batch/load_text.rb

##PS
Generate medical records for evaluating patient similarity (under construction)

##For View
ex)

1. python view/bulk_data_for_parallelcordinates.py output/d_lab.json view/space_parallelCoordinates.html view/done_parallelCoordinates.html
2. python view/bulk_data_for_chart.py view/tchart.json view/space_cumulativeLineChart.html view/done_cumulativeLineChart.html

#Simple Patient History View
1st step:
These codes generate different dataset. Generate Patinets -> python src/json_time_series_generate_patients.py n
n - number of patinets

2nd step:
Generate Graphs -> python src/dump_to_graph.py