Code fragments of Main.py

• # evaluate knn imputation and random forest for the horse colic dataset
• #import math
• import numpy as np
• import pandas as pd
• #import neurolab as nl
• from sklearn.preprocessing import StandardScaler
• from sklearn.svm import SVC
• from missingpy import MissForest
• from minio import Minio
• import os
• def getDataframe(): #ENV Variables
• minio_address = str(os.environ['MINIO_ADDRESS'])
• minio_port = str(os.environ['MINIO_PORT'])
• minio_access_key = str(os.environ['MINIO_ACCESS'])
• minio_secret_key = str(os.environ['MINIO_SECRET'])
• bucket_name = str(os.environ['MINIO_BUCKET_NAME'])
• object_name = str(os.environ['MINIO_OBJECT_NAME'])
• minioClient = Minio( '{0}:{1}'.format(minio_address, minio_port), access_key=minio_access_key, secret_key=minio_secret_key, secure=False, )
• res = minioClient.get_object(bucket_name, object_name)
• print("Data loaded")
• df = pd.read_csv(res)
• return df
• df = getDataframe()
• print("Data loades from MinIO")
• match = lambda a, b: [ b.index(x)+1 if x in b else None for x in a ]
• record_ids = df["record_id"][pd.isna(df["diagnosed_leuk"])==False]
• matched_record_ids = list(match(list(df["record_id"]),list(record_ids)))
• matched_record_ids_none = []
• number = 0
• for x in matched_record_ids:
• if x is not None:
• matched_record_ids_none.append(number)
• number = number + 1
• df_labels = df.iloc[matched_record_ids_none,:]
• df_only_labels = df_labels[df_labels["redcap_repeat_instrument"].isna()]
• rri_list = df_labels["redcap_repeat_instrument"] == "examination_data_use_new_sheet_for_every_visit"
• exam_numbers = []
• number = 0
• for x in rri_list:
• if x is True:
• exam_numbers.append(number)
• number = number + 1
• symptom_df = df_labels.iloc[exam_numbers,:]
• matched_record_ids_labels = match(list(df_only_labels["record_id"]),list(symptom_df["record_id"]))
• matched_record_ids_none_2 = []
• number = 0
• for x in matched_record_ids_labels:
• if x is not None:
• matched_record_ids_none_2.append(number)
• number = number + 1
• labels = list(df_only_labels.iloc[matched_record_ids_none_2,:]["diagnosed_leuk"])
• rri_list_2 = symptom_df["redcap_repeat_instance"] == 1
• exam_numbers = []
• number = 0
• for x in rri_list_2:
• if x is True:
• exam_numbers.append(number)
• number = number + 1
• symptom_first_visit = symptom_df.iloc[exam_numbers,:]
• first_visit_col_number = (symptom_first_visit.columns == "visit1_fir")
• number = 0
• for x in first_visit_col_number:
• if x == True:
• exam_number = number
• number = number + 1
• only_symptoms = symptom_first_visit.iloc[:,exam_number:]
• columns_without_na = []
• for coloums in range(0,len(only_symptoms.iloc[0,:])):
• if not(all(only_symptoms.iloc[:,coloums].isna())):
• columns_without_na.append(coloums)
• only_symptoms_with_out_na = only_symptoms.iloc[:,columns_without_na]
• columns_without_var = []
• for coloums in range(0,len(only_symptoms_with_out_na.iloc[0,:])):
• if (len(set(only_symptoms_with_out_na.iloc[:,coloums])) != 1):
• columns_without_var.append(coloums)
• only_symptoms_final = only_symptoms_with_out_na.iloc[:, columns_without_var]
• only_symptoms_final["visit1_fir"][only_symptoms_final["visit1_fir"].isna()] = -1
• only_symptoms_final["cog"][only_symptoms_final["cog"].isna()] = 0
• try:
• only_symptoms_final["apha"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["apha"].isna())] = -1
• only_symptoms_final["apha"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["apha"].isna()))] = 0
• only_symptoms_final["apha"][only_symptoms_final["apha"].isna()] = 0
• only_symptoms_final["cogloss"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["cogloss"].isna())] = -1
• only_symptoms_final["cogloss"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["cogloss"].isna()))] = 0
• only_symptoms_final["cogloss"][only_symptoms_final["cogloss"].isna()] = 0
• only_symptoms_final["eap"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["eap"].isna())] = -1
• only_symptoms_final["eap"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["eap"].isna()))] = 0
• only_symptoms_final["eap"][only_symptoms_final["eap"].isna()] = 0
• only_symptoms_final["loc"][(only_symptoms_final["cogloss"] == 1) == (only_symptoms_final["eap"].isna())] = -1
• only_symptoms_final["loc"][(list(only_symptoms_final["cogloss"] == 2)) and list((only_symptoms_final["eap"].isna()))] = 0
• only_symptoms_final["loc"][only_symptoms_final["loc"].isna()] = 0
• only_symptoms_final["ic"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["ic"].isna())] = -1
• only_symptoms_final["ic"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["ic"].isna()))] = 0
• only_symptoms_final["ic"][only_symptoms_final["ic"].isna()] = 0
• only_symptoms_final["ii"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["ii"].isna())] = -1
• only_symptoms_final["ii"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["ii"].isna()))] = 0
• only_symptoms_final["ii"][only_symptoms_final["ii"].isna()] = 0
• only_symptoms_final["fati"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["fati"].isna())] = -1
• only_symptoms_final["fati"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["fati"].isna()))] = 0
• only_symptoms_final["fati"][only_symptoms_final["fati"].isna()] = 0
• only_symptoms_final["apr"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["apr"].isna())] = -1
• only_symptoms_final["apr"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["apr"].isna()))] = 0
• only_symptoms_final["apr"][only_symptoms_final["apr"].isna()] = 0
• only_symptoms_final["red_consciousness_confus"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["red_consciousness_confus"].isna())] = -1
• only_symptoms_final["red_consciousness_confus"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["red_consciousness_confus"].isna()))] = 0
• only_symptoms_final["red_consciousness_confus"][only_symptoms_final["red_consciousness_confus"].isna()] = 0
• only_symptoms_final["agnosia"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["agnosia"].isna())] = -1
• only_symptoms_final["agnosia"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["agnosia"].isna()))] = 0
• only_symptoms_final["agnosia"][only_symptoms_final["agnosia"].isna()] = 0
• only_symptoms_final["psychosis"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["psychosis"].isna())] = -1
• only_symptoms_final["psychosis"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["psychosis"].isna()))] = 0
• only_symptoms_final["psychosis"][only_symptoms_final["psychosis"].isna()] = 0
• only_symptoms_final["hallucinations_delusions"][(only_symptoms_final["cog"] == 1) == (only_symptoms_final["hallucinations_delusions"].isna())] = -1
• only_symptoms_final["hallucinations_delusions"][(list(only_symptoms_final["cog"] == 2)) and list((only_symptoms_final["hallucinations_delusions"].isna()))] = 0
• only_symptoms_final["hallucinations_delusions"][only_symptoms_final["hallucinations_delusions"].isna()] = 0
• only_symptoms_final["sleep_disturbance"][only_symptoms_final["sleep_disturbance"].isna()] = -1
• only_symptoms_final["mab"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["mab"].isna())] = -1
• only_symptoms_final["mab"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["mab"].isna()))] = 0
• only_symptoms_final["mab"][only_symptoms_final["mab"].isna()] = 0
• only_symptoms_final["adh"][(only_symptoms_final["mab"] == 1) == (only_symptoms_final["adh"].isna())] = -1
• only_symptoms_final["adh"][(list(only_symptoms_final["mab"] == 2)) and list((only_symptoms_final["adh"].isna()))] = 0
• only_symptoms_final["adh"][only_symptoms_final["adh"].isna()] = 0
• # only_symptoms_final["dbfb"][(only_symptoms_final["mab"] == 1) == (only_symptoms_final["dbfb"].isna())] = -1
• # only_symptoms_final["dbfb"][(list(only_symptoms_final["mab"] == 2)) and list((only_symptoms_final["dbfb"].isna()))] = 0
• # only_symptoms_final["dbfb"][only_symptoms_final["dbfb"].isna()] = 0
• only_symptoms_final["depr"][(only_symptoms_final["mab"] == 1) == (only_symptoms_final["depr"].isna())] = -1
• only_symptoms_final["depr"][(list(only_symptoms_final["mab"] == 2)) and list((only_symptoms_final["depr"].isna()))] = 0
• only_symptoms_final["depr"][only_symptoms_final["depr"].isna()] = 0
• only_symptoms_final["ma"][(only_symptoms_final["mab"] == 1) == (only_symptoms_final["ma"].isna())] = -1
• only_symptoms_final["ma"][(list(only_symptoms_final["mab"] == 2)) and list((only_symptoms_final["ma"].isna()))] = 0
• only_symptoms_final["ma"][only_symptoms_final["ma"].isna()] = 0
• only_symptoms_final["personality"][(only_symptoms_final["mab"] == 1) == (only_symptoms_final["personality"].isna())] = -1
• only_symptoms_final["personality"][(list(only_symptoms_final["mab"] == 2)) and list((only_symptoms_final["personality"].isna()))] = 0
• only_symptoms_final["personality"][only_symptoms_final["personality"].isna()] = 0
• only_symptoms_final["s_e"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["s_e"].isna())] = -1
• only_symptoms_final["s_e"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["s_e"].isna()))] = 0
• only_symptoms_final["s_e"][only_symptoms_final["s_e"].isna()] = 0
• # only_symptoms_final["fs"][(only_symptoms_final["s_e"] == 1) == (only_symptoms_final["fs"].isna())] = -1
• # only_symptoms_final["fs"][(list(only_symptoms_final["s_e"] == 2)) and list((only_symptoms_final["fs"].isna()))] = 0
• # only_symptoms_final["fs"][only_symptoms_final["fs"].isna()] = 0
• only_symptoms_final["fs___2"][(only_symptoms_final["s_e"] == 1) == (only_symptoms_final["fs___2"].isna())] = -1
• only_symptoms_final["fs___2"][(list(only_symptoms_final["s_e"] == 2)) and list((only_symptoms_final["fs___2"].isna()))] = 0
• only_symptoms_final["fs___2"][only_symptoms_final["fs___2"].isna()] = 0
• # only_symptoms_final["gs"][(only_symptoms_final["s_e"] == 1) == (only_symptoms_final["gs"].isna())] = -1
• # only_symptoms_final["gs"][(list(only_symptoms_final["s_e"] == 2)) and list((only_symptoms_final["gs"].isna()))] = 0
• # only_symptoms_final["gs"][only_symptoms_final["gs"].isna()] = 0
• only_symptoms_final["emd"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["emd"].isna())] = -1
• only_symptoms_final["emd"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["emd"].isna()))] = 0
• only_symptoms_final["emd"][only_symptoms_final["emd"].isna()] = 0
• only_symptoms_final["diplopia"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["diplopia"].isna())] = -1
• only_symptoms_final["diplopia"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["diplopia"].isna()))] = 0
• only_symptoms_final["diplopia"][only_symptoms_final["diplopia"].isna()] = 0
• only_symptoms_final["nys"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["nys"].isna())] = -1
• only_symptoms_final["nys"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["nys"].isna()))] = 0
• only_symptoms_final["nys"][only_symptoms_final["nys"].isna()] = 0
• only_symptoms_final["ino"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["ino"].isna())] = -1
• only_symptoms_final["ino"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["ino"].isna()))] = 0
• only_symptoms_final["ino"][only_symptoms_final["ino"].isna()] = 0
• only_symptoms_final["oculomot"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["oculomot"].isna())] = -1
• only_symptoms_final["oculomot"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["oculomot"].isna()))] = 0
• only_symptoms_final["oculomot"][only_symptoms_final["oculomot"].isna()] = 0
• only_symptoms_final["fourth_cranial_nerve_palsy"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["fourth_cranial_nerve_palsy"].isna())] = -1
• only_symptoms_final["fourth_cranial_nerve_palsy"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["fourth_cranial_nerve_palsy"].isna()))] = 0
• only_symptoms_final["fourth_cranial_nerve_palsy"][only_symptoms_final["fourth_cranial_nerve_palsy"].isna()] = 0
• only_symptoms_final["abducens"][(only_symptoms_final["emd"] == 1) == (only_symptoms_final["abducens"].isna())] = -1
• only_symptoms_final["abducens"][(list(only_symptoms_final["emd"] == 2)) and list((only_symptoms_final["abducens"].isna()))] = 0
• only_symptoms_final["abducens"][only_symptoms_final["ino"].isna()] = 0
• only_symptoms_final["thy"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["thy"].isna())] = -1
• only_symptoms_final["thy"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["thy"].isna()))] = 0
• only_symptoms_final["thy"][only_symptoms_final["thy"].isna()] = 0
• only_symptoms_final["fp"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["fp"].isna())] = -1
• only_symptoms_final["fp"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["fp"].isna()))] = 0
• only_symptoms_final["fp"][only_symptoms_final["fp"].isna()] = 0
• only_symptoms_final["od"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["od"].isna())] = -1
• only_symptoms_final["od"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["od"].isna()))] = 0
• only_symptoms_final["od"][only_symptoms_final["od"].isna()] = 0
• only_symptoms_final["hi"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["hi"].isna())] = -1
• only_symptoms_final["hi"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["hi"].isna()))] = 0
• only_symptoms_final["hi"][only_symptoms_final["hi"].isna()] = 0
• only_symptoms_final["hp"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["hp"].isna())] = -1
• only_symptoms_final["hp"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["hp"].isna()))] = 0
• only_symptoms_final["hp"][only_symptoms_final["hp"].isna()] = 0
• only_symptoms_final["trig_neur"][(only_symptoms_final["crn"] == 1) == (only_symptoms_final["trig_neur"].isna())] = -1
• only_symptoms_final["trig_neur"][(list(only_symptoms_final["crn"] == 2)) and list((only_symptoms_final["trig_neur"].isna()))] = 0
• only_symptoms_final["trig_neur"][only_symptoms_final["trig_neur"].isna()] = 0
• only_symptoms_final["spsw"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["spsw"].isna())] = -1
• only_symptoms_final["spsw"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["spsw"].isna()))] = 0
• only_symptoms_final["spsw"][only_symptoms_final["spsw"].isna()] = 0
• only_symptoms_final["dya"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["dya"].isna())] = -1
• only_symptoms_final["dya"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["dya"].isna()))] = 0
• only_symptoms_final["dya"][only_symptoms_final["dya"].isna()] = 0
• only_symptoms_final["scs"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["scs"].isna())] = -1
• only_symptoms_final["scs"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["scs"].isna()))] = 0
• only_symptoms_final["scs"][only_symptoms_final["scs"].isna()] = 0
• only_symptoms_final["dysphon"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["dysphon"].isna())] = -1
• only_symptoms_final["dysphon"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["dysphon"].isna()))] = 0
• only_symptoms_final["dysphon"][only_symptoms_final["dysphon"].isna()] = 0
• only_symptoms_final["slurred_speech"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["slurred_speech"].isna())] = -1
• only_symptoms_final["slurred_speech"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["slurred_speech"].isna()))] = 0
• only_symptoms_final["slurred_speech"][only_symptoms_final["slurred_speech"].isna()] = 0
• only_symptoms_final["bulbar_palsy"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["bulbar_palsy"].isna())] = -1
• only_symptoms_final["bulbar_palsy"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["bulbar_palsy"].isna()))] = 0
• only_symptoms_final["bulbar_palsy"][only_symptoms_final["bulbar_palsy"].isna()] = 0
• only_symptoms_final["pseudobulbar_palsy"][(only_symptoms_final["spsw"] == 1) == (only_symptoms_final["pseudobulbar_palsy"].isna())] = -1
• only_symptoms_final["pseudobulbar_palsy"][(list(only_symptoms_final["spsw"] == 2)) and list((only_symptoms_final["pseudobulbar_palsy"].isna()))] = 0
• only_symptoms_final["pseudobulbar_palsy"][only_symptoms_final["pseudobulbar_palsy"].isna()] = 0
• only_symptoms_final["dyp"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["dyp"].isna())] = -1
• only_symptoms_final["dyp"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["dyp"].isna()))] = 0
• only_symptoms_final["dyp"][only_symptoms_final["dyp"].isna()] = 0
• only_symptoms_final["emp"][(only_symptoms_final["visi"] == 1) == (only_symptoms_final["emp"].isna())] = -1
• only_symptoms_final["emp"][(list(only_symptoms_final["visi"] == 2)) and list((only_symptoms_final["emp"].isna()))] = 0
• only_symptoms_final["emp"][only_symptoms_final["emp"].isna()] = 0
• only_symptoms_final["var"][(only_symptoms_final["visi"] == 1) == (only_symptoms_final["var"].isna())] = -1
• only_symptoms_final["var"][(list(only_symptoms_final["visi"] == 2)) and list((only_symptoms_final["var"].isna()))] = 0
• only_symptoms_final["var"][only_symptoms_final["var"].isna()] = 0
• only_symptoms_final["cvd"][(only_symptoms_final["visi"] == 1) == (only_symptoms_final["cvd"].isna())] = -1
• only_symptoms_final["cvd"][(list(only_symptoms_final["visi"] == 2)) and list((only_symptoms_final["cvd"].isna()))] = 0
• only_symptoms_final["cvd"][only_symptoms_final["cvd"].isna()] = 0
• only_symptoms_final["cvi"][(only_symptoms_final["visi"] == 1) == (only_symptoms_final["cvi"].isna())] = -1
• only_symptoms_final["cvi"][(list(only_symptoms_final["visi"] == 2)) and list((only_symptoms_final["cvi"].isna()))] = 0
• only_symptoms_final["cvi"][only_symptoms_final["cvi"].isna()] = 0
• # only_symptoms_final["visual_field_defect"][(only_symptoms_final["visi"] == 1) == (only_symptoms_final["visual_field_defect"].isna())] = -1
• # only_symptoms_final["visual_field_defect"][(list(only_symptoms_final["visi"] == 2)) and list((only_symptoms_final["visual_field_defect"].isna()))] = 0
• # only_symptoms_final["visual_field_defect"][only_symptoms_final["visual_field_defect"].isna()] = 0
• only_symptoms_final["sim"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["sim"].isna())] = -1
• only_symptoms_final["sim"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["sim"].isna()))] = 0
• only_symptoms_final["sim"][only_symptoms_final["sim"].isna()] = 0
• only_symptoms_final["sper"][(only_symptoms_final["sim"] == 1) == (only_symptoms_final["sper"].isna())] = -1
• only_symptoms_final["sper"][(list(only_symptoms_final["sim"] == 2)) and list((only_symptoms_final["sper"].isna()))] = 0
• only_symptoms_final["sper"][only_symptoms_final["sper"].isna()] = 0
• only_symptoms_final["vs"][(only_symptoms_final["sim"] == 1) == (only_symptoms_final["vs"].isna())] = -1
• only_symptoms_final["vs"][(list(only_symptoms_final["sim"] == 2)) and list((only_symptoms_final["vs"].isna()))] = 0
• only_symptoms_final["vs"][only_symptoms_final["vs"].isna()] = 0
• # only_symptoms_final["sensory_impa"][(only_symptoms_final["sim"] == 1) == (only_symptoms_final["sensory_impa"].isna())] = -1
• # only_symptoms_final["sensory_impa"][(list(only_symptoms_final["sim"] == 2)) and list((only_symptoms_final["sensory_impa"].isna()))] = 0
• # only_symptoms_final["sensory_impa"][only_symptoms_final["sensory_impa"].isna()] = 0
• only_symptoms_final["cersy"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["cersy"].isna())] = -1
• only_symptoms_final["cersy"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["cersy"].isna()))] = 0
• only_symptoms_final["cersy"][only_symptoms_final["cersy"].isna()] = 0
• only_symptoms_final["trem"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["trem"].isna())] = -1
• only_symptoms_final["trem"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["trem"].isna()))] = 0
• only_symptoms_final["trem"][only_symptoms_final["trem"].isna()] = 0
• # only_symptoms_final["hye"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["hye"].isna())] = -1
• # only_symptoms_final["hye"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["hye"].isna()))] = 0
• # only_symptoms_final["hye"][only_symptoms_final["hye"].isna()] = 0
• only_symptoms_final["hyo"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["hyo"].isna())] = -1
• only_symptoms_final["hyo"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["hyo"].isna()))] = 0
• only_symptoms_final["hyo"][only_symptoms_final["hyo"].isna()] = 0
• only_symptoms_final["dyt"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["dyt"].isna())] = -1
• only_symptoms_final["dyt"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["dyt"].isna()))] = 0
• only_symptoms_final["dyt"][only_symptoms_final["dyt"].isna()] = 0
• # only_symptoms_final["dyskin"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["dyskin"].isna())] = -1
• # only_symptoms_final["dyskin"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["dyskin"].isna()))] = 0
• # only_symptoms_final["dyskin"][only_symptoms_final["dyskin"].isna()] = 0
• only_symptoms_final["fmd"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["fmd"].isna())] = -1
• only_symptoms_final["fmd"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["fmd"].isna()))] = 0
• only_symptoms_final["fmd"][only_symptoms_final["fmd"].isna()] = 0
• only_symptoms_final["ataxia"][(only_symptoms_final["cersy"] == 1) == (only_symptoms_final["ataxia"].isna())] = -1
• only_symptoms_final["ataxia"][(list(only_symptoms_final["cersy"] == 2)) and list((only_symptoms_final["ataxia"].isna()))] = 0
• only_symptoms_final["ataxia"][only_symptoms_final["ataxia"].isna()] = 0
• only_symptoms_final["bd"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["bd"].isna())] = -1
• only_symptoms_final["bd"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["bd"].isna()))] = 0
• only_symptoms_final["bd"][only_symptoms_final["bd"].isna()] = 0
• only_symptoms_final["sexd"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["sexd"].isna())] = -1
• only_symptoms_final["sexd"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["sexd"].isna()))] = 0
• only_symptoms_final["sexd"][only_symptoms_final["sexd"].isna()] = 0
• only_symptoms_final["edy"][(only_symptoms_final["sexd"] == 1) == (only_symptoms_final["edy"].isna())] = -1
• only_symptoms_final["edy"][(list(only_symptoms_final["sexd"] == 2)) and list((only_symptoms_final["edy"].isna()))] = 0
• only_symptoms_final["edy"][only_symptoms_final["edy"].isna()] = 0
• only_symptoms_final["ll"][(only_symptoms_final["sexd"] == 1) == (only_symptoms_final["ll"].isna())] = -1
• only_symptoms_final["ll"][(list(only_symptoms_final["sexd"] == 2)) and list((only_symptoms_final["ll"].isna()))] = 0
• only_symptoms_final["ll"][only_symptoms_final["ll"].isna()] = 0
• only_symptoms_final["bi"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["bi"].isna())] = -1
• only_symptoms_final["bi"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["bi"].isna()))] = 0
• only_symptoms_final["bi"][only_symptoms_final["bi"].isna()] = 0
• only_symptoms_final["prs"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["prs"].isna())] = -1
• only_symptoms_final["prs"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["prs"].isna()))] = 0
• only_symptoms_final["prs"][only_symptoms_final["prs"].isna()] = 0
• only_symptoms_final["tprs"][(only_symptoms_final["prs"] == 1) == (only_symptoms_final["tprs"].isna())] = -1
• only_symptoms_final["tprs"][(list(only_symptoms_final["prs"] == 2)) and list((only_symptoms_final["tprs"].isna()))] = 0
• only_symptoms_final["tprs"][only_symptoms_final["tprs"].isna()] = 0
• only_symptoms_final["severity_of_paresis"][(only_symptoms_final["tprs"] == 1) == (only_symptoms_final["severity_of_paresis"].isna())] = -1
• only_symptoms_final["severity_of_paresis"][(list(only_symptoms_final["tprs"] == 2)) and list((only_symptoms_final["severity_of_paresis"].isna()))] = 0
• only_symptoms_final["severity_of_paresis"][only_symptoms_final["severity_of_paresis"].isna()] = 0
• only_symptoms_final["psi"][(only_symptoms_final["prs"] == 1) == (only_symptoms_final["psi"].isna())] = -1
• only_symptoms_final["psi"][(list(only_symptoms_final["prs"] == 2)) and list((only_symptoms_final["psi"].isna()))] = 0
• only_symptoms_final["psi"][only_symptoms_final["psi"].isna()] = 0
• only_symptoms_final["spas"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["spas"].isna())] = -1
• only_symptoms_final["spas"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["spas"].isna()))] = 0
• only_symptoms_final["spas"][only_symptoms_final["spas"].isna()] = 0
• only_symptoms_final["tspas"][(only_symptoms_final["spas"] == 1) == (only_symptoms_final["tspas"].isna())] = -1
• only_symptoms_final["tspas"][(list(only_symptoms_final["spas"] == 2)) and list((only_symptoms_final["tspas"].isna()))] = 0
• only_symptoms_final["tspas"][only_symptoms_final["tspas"].isna()] = 0
• # only_symptoms_final["pai"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["pai"].isna())] = -1
• # only_symptoms_final["pai"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["pai"].isna()))] = 0
• # only_symptoms_final["pai"][only_symptoms_final["pai"].isna()] = 0
• #
• # only_symptoms_final["npa"][(only_symptoms_final["pai"] == 1) == (only_symptoms_final["npa"].isna())] = -1
• # only_symptoms_final["npa"][(list(only_symptoms_final["pai"] == 2)) and list((only_symptoms_final["npa"].isna()))] = 0
• # only_symptoms_final["npa"][only_symptoms_final["npa"].isna()] = 0
• #
• # only_symptoms_final["headache"][(only_symptoms_final["pai"] == 1) == (only_symptoms_final["headache"].isna())] = -1
• # only_symptoms_final["headache"][(list(only_symptoms_final["pai"] == 2)) and list((only_symptoms_final["headache"].isna()))] = 0
• # only_symptoms_final["headache"][only_symptoms_final["headache"].isna()] = 0
• only_symptoms_final["vertigo_dizziness"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["vertigo_dizziness"].isna())] = -1
• only_symptoms_final["vertigo_dizziness"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["vertigo_dizziness"].isna()))] = 0
• only_symptoms_final["vertigo_dizziness"][only_symptoms_final["vertigo_dizziness"].isna()] = 0
• only_symptoms_final["type_of_dizziness"][(only_symptoms_final["vertigo_dizziness"] == 1) == (only_symptoms_final["type_of_dizziness"].isna())] = -1
• only_symptoms_final["type_of_dizziness"][(list(only_symptoms_final["vertigo_dizziness"] == 2)) and list((only_symptoms_final["type_of_dizziness"].isna()))] = 0
• only_symptoms_final["type_of_dizziness"][only_symptoms_final["type_of_dizziness"].isna()] = 0
• only_symptoms_final["gdis"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["gdis"].isna())] = -1
• only_symptoms_final["gdis"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["gdis"].isna()))] = 0
• only_symptoms_final["gdis"][only_symptoms_final["gdis"].isna()] = 0
• # only_symptoms_final["gait_imbalance"][(only_symptoms_final["gdis"] == 1) == (only_symptoms_final["gait_imbalance"].isna())] = -1
• # only_symptoms_final["gait_imbalance"][(list(only_symptoms_final["gdis"] == 2)) and list((only_symptoms_final["gait_imbalance"].isna()))] = 0
• # only_symptoms_final["gait_imbalance"][only_symptoms_final["gait_imbalance"].isna()] = 0
• only_symptoms_final["exgdis"][(only_symptoms_final["gdis"] == 1) == (only_symptoms_final["exgdis"].isna())] = -1
• only_symptoms_final["exgdis"][(list(only_symptoms_final["gdis"] == 2)) and list((only_symptoms_final["exgdis"].isna()))] = 0
• only_symptoms_final["exgdis"][only_symptoms_final["exgdis"].isna()] = 0
• only_symptoms_final["nnsym"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["nnsym"].isna())] = -1
• only_symptoms_final["nnsym"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["nnsym"].isna()))] = 0
• only_symptoms_final["nnsym"][only_symptoms_final["nnsym"].isna()] = 0
• only_symptoms_final["addd"][(only_symptoms_final["nnsym"] == 1) == (only_symptoms_final["addd"].isna())] = -1
• only_symptoms_final["addd"][(list(only_symptoms_final["nnsym"] == 2)) and list((only_symptoms_final["addd"].isna()))] = 0
• only_symptoms_final["addd"][only_symptoms_final["addd"].isna()] = 0
• only_symptoms_final["hypogon"][(only_symptoms_final["nnsym"] == 1) == (only_symptoms_final["hypogon"].isna())] = -1
• only_symptoms_final["hypogon"][(list(only_symptoms_final["nnsym"] == 2)) and list((only_symptoms_final["hypogon"].isna()))] = 0
• only_symptoms_final["hypogon"][only_symptoms_final["hypogon"].isna()] = 0
• only_symptoms_final["pd"][(only_symptoms_final["visit1_fir"] == 1) == (only_symptoms_final["pd"].isna())] = -1
• only_symptoms_final["pd"][(list(only_symptoms_final["visit1_fir"] == 2)) and list((only_symptoms_final["pd"].isna()))] = 0
• only_symptoms_final["pd"][only_symptoms_final["pd"].isna()] = 0
• except: print("Block entry doesnt work")
• only_symptoms_final.drop(columns="visit1_fir",axis=1)
• only_symptoms_final.drop(columns="examination_data_use_new_sheet_for_every_visit_complete",axis=1)
• only_symptoms_final.insert(loc=0, column='label', value=labels)
• print("Data preparation is complete")
• #def quarter(x):
• # return math.ceil(x*4)/4
• """ def ANN(x,y,xt,yt): size = len(x) ########################################## #x = sklearn.preprocessing.normalize(x, norm="l1") #xt = sklearn.preprocessing.normalize([xt], norm="l1") #scaler_x = MinMaxScaler(feature_range=(0, 1)) #x = pd. DataFrame(scaler_x.fit_transform(x)) #xt = pd. DataFrame(scaler_x.fit_transform([xt])) #scaler_y = MinMaxScaler(feature_range=(0, 1)) #y = pd. DataFrame(scaler_y.fit_transform([y])) #yt = pd. DataFrame(scaler_y.fit_transform([[yt]])) maxmin=[] for i in range(0,100): maxmin.append([0, 1]) ########################################## inp = x#.reshape(size,1) tar = y.reshape(size,1) # Create network with 2 layers and random initialized net = nl.net.newff(maxmin,[20, 1]) # Train network error = net.train(inp, tar, epochs=5000, show=100, goal=0.01) # Simulate network out = net.sim(inp) # Plot result #pl.subplot(211) #pl.plot(error) #pl.xlabel('Epoch number') #pl.ylabel('error (default SSE)') #x2 = xt#np.linspace(-6.0,6.0,150) ytt = net.sim([xt]) return ytt ytt=np.round(ytt) yttn=[] for item in ytt: if item[0]==0: yttn.append(0) else: yttn.append(1) return len([a for a in np.isclose(yttn , yt) if(a)]) / len(yttn) * 100 """
• imputer = MissForest(missing_values=-1)
• data_real = only_symptoms_final
• #del data_real[data_real.columns[0]]
• data_imputed = imputer.fit_transform(data_real)
• data = pd.DataFrame(data=data_imputed, columns=data_real.columns.values.tolist())
• print("Data imputation is complete")
• #data = pd.read_csv('./output.csv')
• # normalize dataset with MinMaxScaler
• #scaler = MinMaxScaler(feature_range=(-1, 1))
• #data = pd.DataFrame(scaler.fit_transform(realData))
• linear_result=[]
• rbf_result=[]
• poly_result=[]
• sig_result=[]
• for testIndex in range( len(data)): #data_temp=data #train=data.drop([testIndex]) #test=data.iloc[testIndex]
• data_x = data.iloc[:, 1:]
• y = data.iloc[:, 0]
• #y_class1=y
• """"" y_class1=y.replace(2,1) y_class1=y_class1.replace(103,0) y_class1=y_class1.replace(7,0) y_class1=y_class1.replace(84,0) y_class2=y.replace(2,0) y_class2=y_class2.replace(103,1) y_class2=y_class2.replace(7,0) y_class2=y_class2.replace(84,0) """
• # 2, 84 and 103 realted to LD
• y_class=y.replace(2,0)
• y_class=y_class.replace(103,0)
• y_class=y_class.replace(1,0)
• y_class=y_class.replace(84,1)
• # 7 is related to MS
• y_class=y_class.replace(7,0)
• y_class=y_class.replace(29,0)
• y_class=y_class.replace(60,0)
• """"" y_class4=y.replace(2,0) y_class4=y_class4.replace(103,0) y_class4=y_class4.replace(7,0) y_class4=y_class4.replace(84,1) """
• sc = StandardScaler()
• x_scaled =pd.DataFrame( sc.fit_transform(data_x.values))
• xt_scaled=x_scaled.iloc[testIndex]
• x_scaled=x_scaled.drop([testIndex])
• #y_class1=y_class1.drop([testIndex])
• #y_class2=y_class2.drop([testIndex])
• y_class=y_class.drop([testIndex])
• #y_class4=y_class4.drop([testIndex])
• yt=data.iloc[testIndex].iloc[ 1]
• #y_class1= np.ravel(y_class1)
• #y_class2= np.ravel(y_class2)
• y_class= np.ravel(y_class)
• #y_class4= np.ravel(y_class4)
• #svm_class1 = SVC(kernel='sigmoid', C=1, decision_function_shape='ovo', random_state=0)
• #svm_class2 = SVC(kernel='sigmoid', C=1, decision_function_shape='ovo', random_state=0)
• svm_class_sigmoid = SVC(kernel='sigmoid', C=1, decision_function_shape='ovo', random_state=0,probability=True)
• svm_class_linear = SVC(kernel='linear', C=1, decision_function_shape='ovo', random_state=0,probability=True)
• svm_class_poly = SVC(kernel='poly', C=1, decision_function_shape='ovo', random_state=0,probability=True)
• svm_class_rbf= SVC(kernel='rbf', C=1, decision_function_shape='ovo', random_state=0,probability=True)
• #svm_class4 = SVC(kernel='sigmoid', C=1, decision_function_shape='ovo', random_state=0)
• #svm_class1.fit(x_scaled, y_class1)
• #svm_class2.fit(x_scaled, y_class2)
• svm_class_sigmoid.fit(x_scaled, y_class)
• svm_class_linear.fit(x_scaled, y_class)
• svm_class_poly.fit(x_scaled, y_class)
• svm_class_rbf.fit(x_scaled, y_class)
• #svm_class4.fit(x_scaled, y_class4)
• #y_pre_class1 = svm_class1.predict([xt_scaled])
• #y_pre_class2 = svm_class2.predict([xt_scaled])
• y_pre_class_sigmoid = svm_class_sigmoid.predict([xt_scaled])
• y_pre_class_linear = svm_class_linear.predict([xt_scaled])
• y_pre_class_poly = svm_class_poly.predict([xt_scaled])
• y_pre_class_rbf = svm_class_rbf.predict([xt_scaled])
• #y_pre_class4 = svm_class4.predict([xt_scaled])
• #linear = svm.SVC(kernel='linear', C=1, decision_function_shape='ovo').fit(x, y)
• #rbf = svm.SVC(kernel='rbf', gamma=1, C=1, decision_function_shape='ovo').fit(x, y)
• #poly = svm.SVC(kernel='poly', degree=3, C=1, decision_function_shape='ovo').fit(x, y)
• #sig = svm.SVC(kernel='sigmoid', C=1, decision_function_shape='ovo').fit(x, y)
• #linear_result.append([linear.predict([xt])[0],yt])
• if y_pre_class_sigmoid[0]==1:
• sig_result.append([84,yt])
• else:
• sig_result.append([0,yt])
• if y_pre_class_linear[0]==1:
• linear_result.append([84,yt])
• else:
• linear_result.append([0,yt])
• if y_pre_class_poly[0]==1:
• poly_result.append([84,yt])
• else:
• poly_result.append([0,yt])
• if y_pre_class_rbf[0]==1:
• rbf_result.append([84,yt])
• else:
• rbf_result.append([0,yt])
• # if y_pre_class1[0]==0 and y_pre_class2[0]==1 and y_pre_class3[0]==0 and y_pre_class4[0]==0:
• # SVM_result.append([103,yt])
• # else:
• # if y_pre_class1[0]==0 and y_pre_class2[0]==0 and y_pre_class3[0]==1 and y_pre_class4[0]==0:
• # SVM_result.append([7,yt])
• # else:
• # if y_pre_class1[0]==0 and y_pre_class2[0]==0 and y_pre_class3[0]==0 and y_pre_class4[0]==1:
• # SVM_result.append([84,yt])
• # else:
• # SVM_result.append([0,yt])
• #poly_result.append([poly.predict([xt])[0],yt])
• #sig_result.append([sig.predict([xt])[0],yt])
• #Percent_SVM=0
• #Percent_poly=0
• #Percent_ANN=0
• pd.DataFrame(linear_result).to_csv('binary_linear.csv')
• pd.DataFrame(rbf_result).to_csv('binary_rbf.csv')
• pd.DataFrame(poly_result).to_csv('binary_poly.csv')
• pd.DataFrame(sig_result).to_csv('binary_sig.csv')
• Percent_linear=0
• Percent_poly=0
• Percent_rbf=0
• Percent_sig=0
• for index in range(len(data)):
• if linear_result[index][0]!=linear_result[index][1]:
• Percent_linear=Percent_linear+1
• if rbf_result[index][0]!=rbf_result[index][1]:
• Percent_rbf=Percent_rbf+1
• if poly_result[index][0]!=poly_result[index][1]:
• Percent_poly=Percent_poly+1
• if poly_result[index][0]!=poly_result[index][1]:
• Percent_sig=Percent_sig+1
• Percent_linear=Percent_linear/len(data)
• Percent_poly=Percent_poly/len(data)
• Percent_rbf=Percent_rbf/len(data)
• Percent_sig=Percent_sig/len(data)
• print(Percent_linear)
• print(Percent_poly)
• print(Percent_rbf)
• print(Percent_sig)
• print("Model is finish")
• print("SUCCESS")
• """"" data=pd.DataFrame( data[data.iloc[:, 1]!=84].values) linear_result=[] rbf_result=[] poly_result=[] sig_result=[] #for index in range(len(data)): # if SVM_result[index][84]==SVM_result[index][1]:Percent_SVM=Percent_SVM+1 for testIndex in range( len(data)): train=data.drop([testIndex]) test=data.iloc[testIndex] x = train.iloc[:, 2:].values y = train.iloc[:, 1].values xt = test.iloc[ 2:].values yt = test.iloc[ 1] linear = svm.SVC(kernel='linear', C=1, decision_function_shape='ovo').fit(x, y) rbf = svm.SVC(kernel='rbf', gamma=1, C=1, decision_function_shape='ovo').fit(x, y) poly = svm.SVC(kernel='poly', degree=3, C=1, decision_function_shape='ovo').fit(x, y) sig = svm.SVC(kernel='sigmoid', C=1, decision_function_shape='ovo').fit(x, y) linear_result.append([linear.predict([xt])[0],yt]) rbf_result.append([rbf.predict([xt])[0],yt]) poly_result.append([poly.predict([xt])[0],yt]) sig_result.append([sig.predict([xt])[0],yt]) pd.DataFrame(linear_result).to_csv('multi_class_linear.csv') pd.DataFrame(rbf_result).to_csv('multi_class_rbf.csv') pd.DataFrame(poly_result).to_csv('multi_class_poly.csv') pd.DataFrame(sig_result).to_csv('multi_class_sig.csv') Percent_linear=Percent_SVM/len(data) data.iloc[:,1]=data.iloc[:,1].replace(2,0) data.iloc[:,1]=data.iloc[:,1].replace(103,0.25) data.iloc[:,1]=data.iloc[:,1].replace(7,0.75) data.iloc[:,1]=data.iloc[:,1].replace(84,1) scaler = MinMaxScaler(feature_range=(0, 1)) data = pd.DataFrame(scaler.fit_transform(data)) ANN_result=[] for testIndex in range( len(data)): train=data.drop([testIndex]) test=data.iloc[testIndex] x = train.iloc[:, 2:].values y = train.iloc[:, 1].values xt = test.iloc[ 2:].values yt = test.iloc[ 1] pyt= quarter(ANN(x,y,xt,yt)) ANN_result.append([pyt,yt]) Percent_linear=0 Percent_poly=0 Percent_ANN=0 for index in range(len(data)): if linear_result[index][0]==linear_result[index][1]:Percent_linear=Percent_linear+1 if poly_result[index][0]==poly_result[index][1]:Percent_poly=Percent_poly+1 if ANN_result[index][0]==ANN_result[index][1]:Percent_ANN=Percent_ANN+1 Percent_linear=Percent_linear/len(data) Percent_poly=Percent_poly/len(data) Percent_ANN=Percent_ANN/len(data) #2 103 7 84 class1=data.drop(np.where(data.iloc[:,1] != 2)[0]) class2=data.drop(np.where(data.iloc[:,1] != 103)[0]) class3=data.drop(np.where(data.iloc[:,1] != 7)[0]) class4=data.drop(np.where(data.iloc[:,1] != 84)[0]) """

Graph