atrisovic

import pandas as pd
from huggingface_hub import HfApi
import datetime
import json

df = pd.read_csv('../../data/llm_leaderboard.csv', index_col=1, header=1)

def get_model_info(model_id):
    api = HfApi()
    try:

atrisovic

Step 1: Check analytic data

Is the data you need already on FASSE? Check out the catalog here: https://nsaph.info/analytic.html#analytic-data

If it is not, see step 2.

Step 2: Fill in the form below and add it in the comments here.

The format of the form goes like this:

atrisovic

 |-hospital
 | |-zip.rds
 | |-zip2.rds
 | |-SA_COPD2.Rmd
 | |-review.R
 | |-SA_MI-New.Rmd
 | |-SA_CHF2.Rmd
 | |-SA_LungCancer2.Rmd
 | |-SA_LungCancer-New.Rmd

atrisovic

import pandas as pd
import numpy as np
import json
from simplejson import loads


def get_outcomes():
    """ Get and return ICD codes """""
    f = open('icd_codes.json')
    outcomes_ = json.load(f)

atrisovic

{
    "aki": {
        "icd10": [
            "N17"
        ],
        "icd9": [
            "584"
        ]
    },
    "all_kidney": {

atrisovic

# Before running, activate env:

# export CONDA_ENVS_PATH=/nfs/projects/n/nsaph_common/conda/envs/
# export CONDA_PKGS_PATH=/nfs/projects/n/nsaph_common/conda/pkgs/
# source activate nsaph


## Code to ID hospitalizations

library(data.table)

atrisovic

Get data sample

To get the data sample, we take first 25k rows and last 25k rows from the sample of 59mil rows in bash:

>> tail -n25000 /2016/mbsf_abcd_summary_res000017155_req008183_2016.dat \
        > sample_mbsf_abcd_summary_res000017155_req008183_2016.dat
>> head -n25000 /2016/mbsf_abcd_summary_res000017155_req008183_2016.dat \
        >> sample_mbsf_abcd_summary_res000017155_req008183_2016.dat
# word count:

atrisovic

Rewriting git history for data_requests

What happened

Beneficiery ID numbers were shared in a private GitHub repositry, in the following directories:

data_requests/request_projects/medicaid_duplicate_check_2019_09_27

atrisovic

Dataset stats from Dataverse

in the format DOI, release_year, mime_type

SQL DB query:

SELECT p.authority, p.identifier, f.contenttype, p.publicationdate 
FROM datafile f, dvobject o, dataset s, dvobject p 
WHERE f.id = o.id AND o.owner_id = s.id AND s.id = p.id AND s.harvestingclient_id IS NULL

atrisovic

from flask import Flask, redirect, url_for
from celery import Celery
from celery import Task
from subprocess import PIPE, Popen
import logging, os
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Running locally: