lungs_var3_d8g_222f.py 文件源码

def hu_describe(data, uid, part=""):

    if len(data) == 0:
        nanid = -7777
        d = {   "vol_%s" % part: nanid,
                "min_%s" % part: nanid,
            "max_%s" % part: nanid,
            "mean_%s" % part: nanid,
            "variance_%s" % part: nanid,
            "skewness_%s" % part: nanid,
            "kurtosis_%s" % part:  nanid
            }
    else:
        desc = stats.describe(data, axis=None, nan_policy='omit')  #default policy is 'propagate'
        #names = ["nobs", "min", "max", "mean", "variance", "skewness", "kurtosis"]

        d = {   "vol_%s" % part: desc.nobs,
                "min_%s" % part: desc.minmax[0],
            "max_%s" % part: desc.minmax[1],
            "mean_%s" % part: desc.mean,
            "variance_%s" % part: desc.variance,
            "skewness_%s" % part: desc.skewness,
            "kurtosis_%s" % part:  desc.kurtosis
            }
    #columns = ["id",  "n_volume_%s" % part, "hu_min_%s" % part, "hu_nmax_%s" % part, "hu_mean_%s" % part, "hu_variance_%s" % part,"hu_skewness_%s" % part, "hu_kurtosis_%s" % part]
    #d =       [uid, desc.nobs, desc.minmax[0], desc.minmax[1], desc.mean, desc.variance, desc.skewness, desc.kurtosis]

    #columns = sorted(d.keys())

    df = pd.DataFrame(d, index=[uid]) 

    #df = pd.DataFrame.from_records(d, columns=columns, index=["id"])   

    #df.reset_index(level=0, inplace=True)

    #df.sort_index(axis=1)
    #df.index.name = "id"


    #df = pd.DataFrame.from_dict(d, orient='index')

    return df