GConvLSTM_Simulation_reshape

ITSTGCN

Author

SEOYEON CHOI

Published

May 30, 2023

Simulation Study

Import

import matplotlib.pyplot as plt
import pandas as pd
import numpy as np

Fivenodes

# df5 = pd.read_csv('./simulation_results/2023-06-04_13-29-16.csv')
# df6 = pd.read_csv('./simulation_results/2023-06-04_13-39-55.csv')
# df7 = pd.read_csv('./simulation_results/2023-06-04_13-50-24.csv')
# k=1

df1 = pd.read_csv('./simulation_results/2023-06-07_08-10-42.csv')
df2 = pd.read_csv('./simulation_results/2023-06-07_09-14-08.csv')
df3 = pd.read_csv('./simulation_results/2023-06-07_10-17-23.csv')
df4 = pd.read_csv('./simulation_results/2023-06-07_11-20-35.csv')
df5 = pd.read_csv('./simulation_results/2023-06-07_12-23-45.csv')
df6 = pd.read_csv('./simulation_results/2023-06-07_13-15-54.csv')
df7 = pd.read_csv('./simulation_results/2023-06-13_11-25-39.csv')
df8 = pd.read_csv('./simulation_results/2023-06-13_12-25-51.csv')

df9 = pd.read_csv('./simulation_results/2023-07-16_09-20-23.csv')
df10 = pd.read_csv('./simulation_results/2023-07-16_10-22-53.csv')
df11 = pd.read_csv('./simulation_results/2023-07-16_11-27-39.csv')
df12 = pd.read_csv('./simulation_results/2023-07-16_12-30-05.csv')
df13 = pd.read_csv('./simulation_results/2023-07-16_13-33-28.csv')
df14 = pd.read_csv('./simulation_results/2023-07-16_14-34-12.csv')

data = pd.concat([df1,df2,df3,df4,df5,df6,df7,df8,df9,df10,df11,df12,df13,df14],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_fivedones_Simulation.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_fivedones_Simulation.csv')

Baseline

data.query("method!='GNAR' and mrate==0").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='nof_filters',height=600)

Random

data.query("method!='GNAR' and mtype =='rand'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='nof_filters',height=600)

Block

data.query("method!='GNAR' and mtype =='block' ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='nof_filters',height=600)

chickenpox

# df1 = pd.read_csv('./simulation_results/2023-06-02_14-03-24.csv') # GNAR random
# df2 = pd.read_csv('./simulation_results/2023-06-03_15-57-09.csv') # GNAR block
# df3 = pd.read_csv('./simulation_results/2023-06-03_17-58-22.csv') # STGCN, ITSTGCN random 30%

## 하이퍼 파라메터 조정해서 다시 함!
df1 = pd.read_csv('./simulation_results/2023-06-07_22-08-57.csv') 
df2 = pd.read_csv('./simulation_results/2023-06-08_01-57-25.csv') 
df3 = pd.read_csv('./simulation_results/2023-06-08_03-58-56.csv') 
df4 = pd.read_csv('./simulation_results/2023-06-08_06-00-50.csv')
df5 = pd.read_csv('./simulation_results/2023-06-08_09-11-04.csv')
df6 = pd.read_csv('./simulation_results/2023-06-08_13-06-18.csv')

df7 = pd.read_csv('./simulation_results/2023-07-16_18-29-09.csv')
df8 = pd.read_csv('./simulation_results/2023-07-16_22-21-02.csv')
df9 = pd.read_csv('./simulation_results/2023-07-17_02-20-42.csv')
df10 = pd.read_csv('./simulation_results/2023-07-17_06-23-41.csv')

data = pd.concat([df1,df2,df3,df4,df5,df6,df7,df8,df9,df10],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_chikenpox_Simulation.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_chikenpox_Simulation.csv')

Baseline

data.query("method!='GNAR' and mrate ==0 ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='nof_filters',height=600)

Random

data.query("method!='GNAR' and mtype =='rand'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='nof_filters',facet_row='inter_method',height=800)

Block

data.query("method!='GNAR' and mtype =='block'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='lags',height=800)

Pedalme

# df3 = pd.read_csv('./simulation_results/2023-06-04_13-03-18.csv') 
# df4 = pd.read_csv('./simulation_results/2023-06-04_13-19-20.csv') 
# k=1

df1 = pd.read_csv('./simulation_results/2023-06-08_14-08-24.csv') 
df2 = pd.read_csv('./simulation_results/2023-06-10_14-21-35.csv')

df3 = pd.read_csv('./simulation_results/2023-07-17_07-05-13.csv')

data = pd.concat([df1,df2,df3],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_pedalme_Simulation.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_pedalme_Simulation.csv')

Baseline

data.query("mrate ==0 and lags!=2").plot.box(backend='plotly',x='epoch',color='method',y='mse',facet_col='nof_filters',facet_row='lags',height=400)

random

data.query("method!='GNAR' and mtype =='rand' and lags!=2").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='nof_filters',facet_row='inter_method',height=800)

block

data.query("method!='GNAR' and mtype =='block' and lags!=2 ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='nof_filters',facet_row='inter_method',height=800)

weight matrix time, node 고려한 결과

df1 = pd.read_csv('./simulation_results/2023-06-10_15-33-00.csv')
df2 = pd.read_csv('./simulation_results/2023-06-10_16-06-20.csv')

data2 = pd.concat([df1,df2],axis=0)

data2.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_pedalme_Simulation_itstgcnsnd.csv',index=False)

data2 = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_pedalme_Simulation_itstgcnsnd.csv')

data2.query("mtype=='rand'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='nof_filters',height=1000)

data2.query("mtype=='block'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='nof_filters',height=1000)

Wikimath

df1 = pd.read_csv('./simulation_results/2023-06-09_05-04-53.csv') # 0.3
df2 = pd.read_csv('./simulation_results/2023-06-09_22-51-24.csv') # 0.3
df3 = pd.read_csv('./simulation_results/2023-06-10_13-44-50.csv') # 0.3
df4 = pd.read_csv('./simulation_results/2023-06-11_17-52-38.csv') # 0.8
df5 = pd.read_csv('./simulation_results/2023-06-12_22-28-48.csv') # 0.8
df6 = pd.read_csv('./simulation_results/2023-06-13_19-20-20.csv') # 0
df7 = pd.read_csv('./simulation_results/2023-06-14_20-51-43.csv') # 0
df8 = pd.read_csv('./simulation_results/2023-06-15_17-05-25.csv') # 0
df9 = pd.read_csv('./simulation_results/2023-06-13_16-54-57.csv') # block
df10 = pd.read_csv('./simulation_results/2023-06-12_20-55-27.csv') # block
df11 = pd.read_csv('./simulation_results/2023-06-12_01-29-40.csv') # block
# 11개 끝나면 같은 노드 missing 으로 넘어가기

df12 = pd.read_csv('./simulation_results/2023-07-22_22-49-59.csv') # 0.3
df13 = pd.read_csv('./simulation_results/2023-07-23_13-08-36.csv') # 0.3
df14 = pd.read_csv('./simulation_results/2023-07-24_03-55-14.csv') # 0.3
df15 = pd.read_csv('./simulation_results/2023-07-24_16-16-52.csv') # 0.8
df16 = pd.read_csv('./simulation_results/2023-07-25_05-43-47.csv') # 0.8
df17 = pd.read_csv('./simulation_results/2023-07-25_16-41-57.csv') # 0
df18 = pd.read_csv('./simulation_results/2023-07-26_05-16-34.csv') # 0
df19 = pd.read_csv('./simulation_results/2023-07-26_17-46-04.csv') # 0
df20 = pd.read_csv('./simulation_results/2023-07-27_06-03-29.csv') # block
df21 = pd.read_csv('./simulation_results/2023-07-27_16-41-32.csv') # block
df22 = pd.read_csv('./simulation_results/2023-07-28_03-28-10.csv') 
df23 = pd.read_csv('./simulation_results/2023-07-28_14-03-06.csv')

data = pd.concat([df1,df2,df3,df4,df5,df6,df7,df8,df9,df10,df11,df12,df13,df14,df15,df16,df17,df18,df19,df20,df21,df22,df23],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_wikimath.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_wikimath.csv')

Baseline

data.query("mrate==0 ").plot.box(backend='plotly',x='mrate',color='method',y='mse',height=600)

random

data.query("mtype=='rand' and mrate !=0 and method!='GNAR'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='nof_filters',height=800)

block

data.query("mtype=='block' ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='nof_filters',height=1200)

missing values on the same nodes

df1 = pd.read_csv('./simulation_results/2023-06-14_19-23-44.csv') # STGCN IT-STGCN block
df2 = pd.read_csv('./simulation_results/2023-06-15_15-51-38.csv')
df3 = pd.read_csv('./simulation_results/2023-06-16_04-32-51.csv')

data = pd.concat([df1,df2,df3],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_wikimath_GSO_st.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_wikimath_GSO_st.csv')

data.query("method!='GNAR' and mtype =='block' ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='inter_method',facet_row='lags',height=800)

Windmilsmall

df1 = pd.read_csv('./simulation_results/2023-06-18_08-49-16.csv') # STGCN IT-STGCN 70%
df2 = pd.read_csv('./simulation_results/2023-06-18_19-23-41.csv') # STGCN IT-STGCN
df3 = pd.read_csv('./simulation_results/2023-06-19_05-43-03.csv') # STGCN IT-STGCN
# df4 = pd.read_csv('./simulation_results/2023-06-19_05-43-03.csv') # STGCN IT-STGCN
df5 = pd.read_csv('./simulation_results/2023-06-19_15-49-47.csv') # STGCN IT-STGCN
df6 = pd.read_csv('./simulation_results/2023-06-20_01-48-58.csv') #
df7 = pd.read_csv('./simulation_results/2023-06-20_10-46-07.csv') # STGCN IT-STGCN
df8 = pd.read_csv('./simulation_results/2023-06-20_22-24-25.csv') # STGCN IT-STGCN
df9 = pd.read_csv('./simulation_results/2023-06-21_08-49-11.csv') # STGCN IT-STGCN
df10 = pd.read_csv('./simulation_results/2023-06-21_18-48-04.csv') # STGCN IT-STGCN
df11 = pd.read_csv('./simulation_results/2023-06-22_04-57-53.csv') # STGCN IT-STGCN
df12 = pd.read_csv('./simulation_results/2023-06-23_01-51-16.csv') # STGCN IT-STGCN
df13 = pd.read_csv('./simulation_results/2023-06-23_09-45-30.csv') # STGCN IT-STGCN
df14 = pd.read_csv('./simulation_results/2023-06-23_17-38-53.csv') # STGCN IT-STGCN
df15 = pd.read_csv('./simulation_results/2023-06-24_03-32-00.csv') # STGCN IT-STGCN
df16 = pd.read_csv('./simulation_results/2023-06-24_14-02-04.csv') # STGCN IT-STGCN
df17 = pd.read_csv('./simulation_results/2023-06-25_00-33-49.csv') # STGCN IT-STGCN
df18 = pd.read_csv('./simulation_results/2023-06-25_11-14-03.csv') # STGCN IT-STGCN
df19 = pd.read_csv('./simulation_results/2023-06-25_21-52-31.csv') # STGCN IT-STGCN
df20 = pd.read_csv('./simulation_results/2023-06-26_09-32-33.csv') # STGCN IT-STGCN
df21 = pd.read_csv('./simulation_results/2023-06-26_20-26-47.csv') # STGCN IT-STGCN
df22 = pd.read_csv('./simulation_results/2023-06-30_22-22-17.csv') # STGCN IT-STGCN
df23 = pd.read_csv('./simulation_results/2023-06-27_07-38-02.csv') # STGCN IT-STGCN
df24 = pd.read_csv('./simulation_results/2023-06-27_17-42-21.csv') # STGCN IT-STGCN
df25 = pd.read_csv('./simulation_results/2023-06-27_23-25-36.csv') # STGCN IT-STGCN
df26 = pd.read_csv('./simulation_results/2023-06-28_04-51-32.csv') # STGCN IT-STGCN
df27 = pd.read_csv('./simulation_results/2023-06-28_13-42-03.csv') # STGCN IT-STGCN
df28 = pd.read_csv('./simulation_results/2023-06-28_23-31-39.csv') # STGCN IT-STGCN
df29 = pd.read_csv('./simulation_results/2023-06-29_07-37-23.csv') # STGCN IT-STGCN
df30 = pd.read_csv('./simulation_results/2023-06-29_11-42-51.csv') # STGCN IT-STGCN
df31 = pd.read_csv('./simulation_results/2023-06-29_18-56-11.csv') # GNAR 70%
df32 = pd.read_csv('./simulation_results/2023-06-29_23-00-06.csv') # GNAR 
df33 = pd.read_csv('./simulation_results/2023-06-30_07-55-33.csv')
df34 = pd.read_csv('./simulation_results/2023-06-30_13-16-45.csv')
df35 = pd.read_csv('./simulation_results/2023-06-30_22-22-17.csv')
df36 = pd.read_csv('./simulation_results/2023-07-01_02-32-38.csv')
df37 = pd.read_csv('./simulation_results/2023-05-02_01-10-53.csv')
df38 = pd.read_csv('./simulation_results/2023-07-01_22-06-17.csv')
df39 = pd.read_csv('./simulation_results/2023-07-02_02-39-51.csv')
df40 = pd.read_csv('./simulation_results/2023-07-02_10-30-04.csv')
df41 = pd.read_csv('./simulation_results/2023-07-02_14-55-00.csv')
df42 = pd.read_csv('./simulation_results/2023-07-02_22-39-16.csv')
df43 = pd.read_csv('./simulation_results/2023-07-03_02-31-23.csv')
df44 = pd.read_csv('./simulation_results/2023-07-03_09-46-01.csv')
df45 = pd.read_csv('./simulation_results/2023-07-03_13-39-49.csv')
df46 = pd.read_csv('./simulation_results/2023-07-03_20-27-34.csv')
df47 = pd.read_csv('./simulation_results/2023-07-03_23-24-36.csv')
df48 = pd.read_csv('./simulation_results/2023-07-04_04-53-17.csv')
df49 = pd.read_csv('./simulation_results/2023-07-04_07-47-18.csv')
df50 = pd.read_csv('./simulation_results/2023-07-04_13-19-43.csv')
df51 = pd.read_csv('./simulation_results/2023-07-04_16-15-28.csv')
df52 = pd.read_csv('./simulation_results/2023-07-04_21-45-58.csv')
df53 = pd.read_csv('./simulation_results/2023-07-05_00-42-02.csv')
df54 = pd.read_csv('./simulation_results/2023-07-05_06-10-53.csv')
df55 = pd.read_csv('./simulation_results/2023-07-05_09-04-03.csv')
df56 = pd.read_csv('./simulation_results/2023-07-05_14-35-14.csv')
df57 = pd.read_csv('./simulation_results/2023-07-05_17-33-38.csv')
df58 = pd.read_csv('./simulation_results/2023-07-05_23-06-14.csv')
df59 = pd.read_csv('./simulation_results/2023-07-06_01-59-24.csv')
df60 = pd.read_csv('./simulation_results/2023-07-06_07-30-43.csv')
df61 = pd.read_csv('./simulation_results/2023-07-06_10-26-11.csv')
df62 = pd.read_csv('./simulation_results/2023-07-06_15-56-04.csv')
df63 = pd.read_csv('./simulation_results/2023-07-06_18-51-39.csv')
df64 = pd.read_csv('./simulation_results/2023-07-07_00-20-44.csv')
df65 = pd.read_csv('./simulation_results/2023-07-07_03-16-52.csv')
df66 = pd.read_csv('./simulation_results/2023-07-07_08-45-18.csv')
df67 = pd.read_csv('./simulation_results/2023-07-07_11-42-32.csv')
df68 = pd.read_csv('./simulation_results/2023-07-07_17-11-59.csv')
df69 = pd.read_csv('./simulation_results/2023-07-07_20-09-16.csv')
df70 = pd.read_csv('./simulation_results/2023-07-08_01-39-33.csv') 
df71 = pd.read_csv('./simulation_results/2023-07-08_04-35-48.csv')
df72 = pd.read_csv('./simulation_results/2023-07-08_10-41-29.csv')
df73 = pd.read_csv('./simulation_results/2023-07-08_14-48-52.csv')
df74 = pd.read_csv('./simulation_results/2023-07-08_22-15-53.csv')
df75 = pd.read_csv('./simulation_results/2023-07-09_02-11-41.csv')
df76 = pd.read_csv('./simulation_results/2023-07-09_09-59-35.csv')
df77 = pd.read_csv('./simulation_results/2023-07-09_14-17-00.csv')
df78 = pd.read_csv('./simulation_results/2023-07-09_22-10-25.csv')
df79 = pd.read_csv('./simulation_results/2023-07-10_02-39-25.csv')
df80 = pd.read_csv('./simulation_results/2023-07-10_12-33-23.csv')
df81 = pd.read_csv('./simulation_results/2023-07-10_17-03-55.csv')
df82 = pd.read_csv('./simulation_results/2023-07-11_01-43-12.csv')
df83 = pd.read_csv('./simulation_results/2023-07-11_05-19-38.csv')
df84 = pd.read_csv('./simulation_results/2023-07-11_13-21-44.csv')
df85 = pd.read_csv('./simulation_results/2023-07-11_16-56-36.csv')
df86 = pd.read_csv('./simulation_results/2023-07-12_05-06-36.csv')
df87 = pd.read_csv('./simulation_results/2023-07-12_10-10-49.csv')
df88 = pd.read_csv('./simulation_results/2023-07-12_17-03-55.csv')
df89 = pd.read_csv('./simulation_results/2023-07-13_15-31-01.csv')
df90 = pd.read_csv('./simulation_results/2023-07-14_20-29-43.csv')
df91 = pd.read_csv('./simulation_results/2023-07-13_04-24-40.csv')
df92 = pd.read_csv('./simulation_results/2023-07-15_09-33-11.csv')

data = pd.concat([df1,df2,df3,df5,df6,df7,df8,df9,df10,df11,df12,df13,df14,df15,df16,df17,df18,df19,df20,df21,
                 df22,df23,df24,df25,df26,df27,df28,df29,df30,df31,df32,df33,df34,df35,df36,df37,df38,df39,df40,
                 df41,df42,df43,df44,df45,df46,df47,df48,df29,df50,df51,df52,df53,df54,df55,df56,df57,df58,df59,df60,df61,
                 df62,df63,df64,df65,df66,df67,df68,df69,df70,df71,df72,df73,df74,df75,df76,df77,df78,df79,df80,df81,
                 df82,df83,df84,df85,df86,df87,df88,df89,df90,df91,df92],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_windmillsmall.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_windmillsmall.csv')

Baseline

data.query("method!='GNAR' and mrate ==0 ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',height=800)

random

data.query("method!='GNAR' and mtype =='rand' ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='inter_method',height=800)

block

data.query("method!='GNAR' and mtype =='block' ").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='inter_method',height=800)

Montevideobus

df1 = pd.read_csv('./simulation_results/2023-06-09_04-44-47.csv') # lags 8
df2 = pd.read_csv('./simulation_results/2023-06-09_10-14-28.csv') # lags 8
df3 = pd.read_csv('./simulation_results/2023-06-09_15-02-44.csv') # lags 8
df4 = pd.read_csv('./simulation_results/2023-06-09_20-18-35.csv') # lags 8
df5 = pd.read_csv('./simulation_results/2023-06-10_19-55-08.csv') # lags 8
df6 = pd.read_csv('./simulation_results/2023-06-10_23-03-00.csv') # lags 8
df7 = pd.read_csv('./simulation_results/2023-06-11_02-22-49.csv') # lags 8

df8 = pd.read_csv('./simulation_results/2023-07-20_01-06-37.csv') # lags 8
df9 = pd.read_csv('./simulation_results/2023-07-20_06-05-46.csv') # lags 8
df10 = pd.read_csv('./simulation_results/2023-07-20_10-50-04.csv') # lags 8
df11 = pd.read_csv('./simulation_results/2023-07-20_15-35-45.csv') # lags 8
df12 = pd.read_csv('./simulation_results/2023-07-20_20-12-53.csv') # lags 8
df13 = pd.read_csv('./simulation_results/2023-07-21_00-58-20.csv') # lags 8
df14 = pd.read_csv('./simulation_results/2023-07-21_05-45-10.csv') # lags 8
df15 = pd.read_csv('./simulation_results/2023-07-21_10-13-49.csv') # lags 8
df16 = pd.read_csv('./simulation_results/2023-07-21_14-15-40.csv') # lags 8

data = pd.concat([df1,df2,df3,df4,df5,df6,df7,df8,df9,df10,df11,df12,df13,df14,df15,df16],axis=0)

data.to_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_monte.csv',index=False)

data = pd.read_csv('./simulation_results/Real_simulation_reshape/GConvLSTM_monte.csv')

Baseline

data.query("mrate==0 and method!='GNAR'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='nof_filters',height=600)

random

data.query("mtype=='rand' and mrate !=0 and method!='GNAR'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='inter_method',height=800)

block

data.query("mtype=='block' and method!='GNAR'").plot.box(backend='plotly',x='mrate',color='method',y='mse',facet_col='lags',facet_row='nof_filters',height=1200)