[WIP] convert HBN ASR

[yibeichan]

I converted ASR (one questionnaire from HBN) as my first try for reproschema. It would be great if @sooyounga or @djarecka can take a look at it (especially items/ASR_126). (btw, @djarecka I went through LinkML tutorials but think it's easier to write hard-coded python script for this first converting since I'm not that familiar with schema yet. )

Also, based on my understanding of this questionnaire, our final output should be structured similar to this one, where items should be grouped into sub-categories (e.g., ASR_SC, ASR_AB, ASR_WD, etc.). I don't have detailed information about which items should be group which category at this moment but I'll update the structure once I have more info on the questionnaire.

[yibeichan]

for reference, i put my python script here (with directory info removed):

df = pd.read_csv('hbn/item-names-cleaned.csv')

# choose one assessment
asr_df = df[df["datadic"]=="ASR"]
asr_df.dropna(subset=["keys"], inplace=True)
sub_asr_df = asr_df[asr_df["keys"].str.match(r'^ASR_\d+[a-z]?$')].copy()
# fill nan as 999
sub_asr_df["value"].fillna(999, inplace=True)

# create json file for each item
for index, row in sub_asr_df.iterrows():
    var = row["keys"]
    dom = row["domains"].replace('_', ' ')
    desc = f"Q{index+1} of the {dom}"
    question = row["questions"].replace(r'^\d+\.\s', '')
    if row["value"] != 999:
        itype = "radio"
        resopt = "../valueConstraints"
    else:
        itype = "text"
        resopt = {
            "valueType": "xsd:string"
            }
    item_json = {
        "@context": "https://raw.githubusercontent.com/ReproNim/reproschema/1.0.0-rc4/contexts/generic",
        "@type": "reproschema:Field",
        "@id": var,
        "prefLabel": var,
        "description": desc,
        "schemaVersion": "1.0.0-rc4",
        "version": "0.0.1",
        "question": {
            "en": question
        },
        "ui": {
            "inputType": itype
        },
        "responseOptions": resopt
    }
    with open(os.path.join(items_dir, f"{var}"), 'w', encoding='utf-8') as jsonfile:
        json.dump(item_json, jsonfile, indent=4)

# create the main json file
vars = sub_asr_df["keys"]
# Build the JSON-LD object for the main file
jsonld_obj = {
    "@context": "https://raw.githubusercontent.com/ReproNim/reproschema/1.0.0-rc1/contexts/generic",
    "@type": "reproschema:Activity",
    "@id": "hbn_asr",
    "prefLabel": "HBN Adult Self Report (ASR)",
    "altLabel": "hbn_asr",
    "schemaVersion": "1.0.0-rc1",
    "version": "0.0.1",
    "ui": {
        "order": [f"items/{var}" for var in vars],
        "shuffle": False,
        "addProperties": [{"variableName": var, "isAbout": f"items/{var}"} for var in vars]
    }
}

# Write the JSON-LD object to a file
with open(os.path.join(output_dir, "hbn_asr_schema"), 'w', encoding='utf-8') as jsonfile:
    json.dump(jsonld_obj, jsonfile, indent=4)

[djarecka]

item-names-cleaned.csv is somewhere online?

[yibeichan]

@djarecka can you run the workflow again?