Skip to content

GitLab

Commit 686e249c authored by Leonie Pick's avatar Leonie Pick
Browse files
Options

Finished review work on code

parent da98c4d7
Expand all Show whitespace changes
Inline Side-by-side
Showing with 891 additions and 193 deletions
.gitignore
View file @ 686e249c
......@@ -3,6 +3,7 @@ Dump/
# Don't track the following files
Input.nc
Input_old.nc
Readme.md
# Don't track all files with these endings:
......
ClassifyStorms.ipynb
View file @ 686e249c
This diff is collapsed.
Modules.py
View file @ 686e249c
......@@ -55,7 +55,7 @@ def Search_TargetEvents(HMC, HMC11y, HMC5d, dHMC, HTime, DTime, grid, Plot, Save
elif (IndexDiff[i] > 1): # difference is greater than 1
if j >= jmax:
jmax= j
jmax = j
if j > mini_range:
......@@ -135,7 +135,7 @@ def Search_TargetEvents(HMC, HMC11y, HMC5d, dHMC, HTime, DTime, grid, Plot, Save
return Index_thres1, StormIndices, IndexMin1, IndexMin2
###
###
def Get_TargetEvents(HMC, HMC11y, HMC5d, dHMC, Kp_all, KpHours_all, Reference, Time, Date, YearStart, Plot, Save):
def Get_TargetEvents(HMC, HMC11y, HMC5d, dHMC, Kp_all, KpHours_all, Dst, SYM, SYMMinutes, Reference, Time, Date, YearStart, Mode, Plot, Save):
YearsIndex = np.where(Time[:,0] >= YearStart)[0]
TrClass = Reference[:,0]
......@@ -172,7 +172,7 @@ def Get_TargetEvents(HMC, HMC11y, HMC5d, dHMC, Kp_all, KpHours_all, Reference, T
Time[YearsIndex,:], Date[YearsIndex], grid[i,:], Plot, Save)
if Plot == True:
pl.Selection(Time[YearsIndex,:], Date[YearsIndex], HMC[YearsIndex], Index_thres, Storms_a, Storms_b, Storms, Kp_all, KpHours_all, Save)
pl.Selection(Time[YearsIndex,:], Date[YearsIndex], HMC[YearsIndex], Index_thres, Storms_a, Storms_b, Storms, Kp_all, KpHours_all, Dst[YearsIndex], SYM, SYMMinutes, Save)
#Found_CIRs = np.where(np.in1d(TrTimeIndex[np.logical_or(TrClass==0,TrClass==2)],Storms+YearsIndex[0]))[0]
#Found_CMEs = np.where(np.in1d(TrTimeIndex[TrClass==1],Storms+YearsIndex[0]))[0]
......@@ -208,14 +208,14 @@ def Get_TargetEvents(HMC, HMC11y, HMC5d, dHMC, Kp_all, KpHours_all, Reference, T
###
TargetResults = {'No. target events': len(StormsWin[SelectStr]),'Fraction [%]':len(StormsWin[SelectStr])*100/len(HMC[YearsIndex])}
TrainingResults = [[len(CIRs), len(CMEs), len(CIRs)+len(CMEs)],[len(FoundStorms['CIRs_'+SelectStr]),len(FoundStorms['CMEs_'+SelectStr]),len(FoundStorms['CIRs_'+SelectStr])+len(FoundStorms['CMEs_'+SelectStr])]]
SelectionParams = {r'$P_n [nT]$':grid[:,0],r'$p_{sc}$':grid[:,1], r'$\Delta t$ [h]':grid[:,2], r' $Hs$ [nT]':-grid[:,3]}
SelectionParams = {r'$n$':grid[:,0],r'$p_{sc}$':grid[:,1], r'$\Delta t$ [h]':grid[:,2]+1, r' $Hs$ [nT]':-grid[:,3]}
display(pd.DataFrame(data=SelectionParams))
display(pd.DataFrame(data=TargetResults,index=['Selection']))
display(pd.DataFrame(data=TrainingResults,columns=['No. CIRs','No. CMEs', 'Total'], index=['Reference set', 'Training set']))
###
if Plot == True:
pl.IndexDist(Time,YearsIndex,StormsWin[SelectStr],TrTimeIndex,Kp_all,KpHours_all,HMC,Save)
pl.IndexDist(Time,YearsIndex,StormsWin[SelectStr],TrTimeIndex,Kp_all,KpHours_all,HMC,Dst,Mode,Save)
return StormsWin, TrFound, FoundStorms
###
......
Plots.py
View file @ 686e249c
from Imports import *
###
def Selection(Time,Date,HMC,Index_thres1,StormIndices,IndexMin1,IndexMin2,Kp,KpHours,Save):
def Selection(Time,Date,HMC,Index_thres1,StormIndices,IndexMin1,IndexMin2,Kp,KpHours,Dst,SYM,SYMMinutes,Save):
fig, ax = plt.subplots(4,1,facecolor='w', edgecolor='k')
fig.set_size_inches([6.69, 8.86], forward=True)
......@@ -19,25 +19,36 @@ def Selection(Time,Date,HMC,Index_thres1,StormIndices,IndexMin1,IndexMin2,Kp,KpH
dot_size=20
# HMC
ax[0].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,label='HMC',zorder=0)
ax[1].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,zorder=0)
ax[2].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,zorder=1)
ax[0].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,label='HMC',zorder=1)
ax[1].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,zorder=1)
ax[2].plot(Time[start:end,4],HMC[start:end],color='black',linewidth=0.75,zorder=2)
# HMC selection
ax[0].plot(Time[start:end,4],Index_thres1[start:end],linestyle='--',color='black',label=r'$Hl_t\in$ ['+str(np.around(max(Index_thres1[start:end]),2))+' nT, '+str(np.around(min(Index_thres1[start:end]),2))+' nT]',zorder=2)
ax[0].plot(Time[start:end,4],Index_thres1[start:end],linewidth=0.75,linestyle='--',color='black',label=r'$Hl_t\in$ ['+str(np.around(max(Index_thres1[start:end]),2))+' nT, '+str(np.around(min(Index_thres1[start:end]),2))+' nT]',zorder=2)
ax[0].scatter(Time[StormIndices,4],HMC[StormIndices],color='black',s=10,zorder=4)
ax[1].scatter(Time[np.asarray(IndexMin1,dtype=int),4],HMC[np.asarray(IndexMin1,dtype=int)],color='black',s=dot_size,zorder=2)
ax[2].scatter(Time[np.asarray(IndexMin2,dtype=int),4],HMC[np.asarray(IndexMin2,dtype=int)],color='black',s=dot_size,zorder=3)
# Dst
Dst_color='dimgray'
ax[0].plot(Time[start:end,4],Dst[start:end],color=Dst_color,linewidth=0.75,label='Dst',zorder=0)
ax[1].plot(Time[start:end,4],Dst[start:end],color=Dst_color,linewidth=0.75,zorder=0)
ax[2].plot(Time[start:end,4],Dst[start:end],color=Dst_color,linewidth=0.75,zorder=1)
# Dst storm scale
G1Dst= Rectangle((Time[start,4],-30),len(Time[start:end,4]),-20,edgecolor='yellow',facecolor='yellow',zorder=0,alpha=0.3,label='weak')
G2Dst = Rectangle((Time[start,4],-50),len(Time[start:end,4]),-50,edgecolor='gold',facecolor='gold',zorder=0,alpha=0.3,label='moderate')
#G3Dst = Rectangle((Time[start,4],-100),len(Time[start:end,4]),-100,edgecolor='orange',facecolor='orange',zorder=0,alpha=0.3,label='strong')
#G4Dst = Rectangle((Time[start,4],-200),len(Time[start:end,4]),-150,edgecolor='red',facecolor='red',zorder=0,alpha=0.3,label='severe')
#G5Dst = Rectangle((Time[start,4],-300),len(Time[start:end,4]),-600,edgecolor='maroon',facecolor='maroon',zorder=0,alpha=0.3,label='great')
ax[2].add_patch(G1Dst);ax[2].add_patch(G2Dst)#;ax[2].add_patch(G3Dst);ax[2].add_patch(G4Dst);ax[2].add_patch(G5Dst)
# Kp
startKPT = 1981.665925; endKPT = 1981.747774
startKP = np.where(KpHours==startKPT)[0][0]
endKP = np.where(KpHours==endKPT)[0][0]
# Kp
# Kp storm scale
Kp_step = 0.333333
ax3b = ax[3].twinx()
ax3b.bar(Time[start:end,4][::3],Kp[startKP:endKP+1],3/(24*365),align='edge',edgecolor=Dst_color,color=Dst_color,zorder=1)
......@@ -51,6 +62,7 @@ def Selection(Time,Date,HMC,Index_thres1,StormIndices,IndexMin1,IndexMin2,Kp,KpH
ax[3].patch.set_visible(False)
# Sym-H
ax[3].plot(SYMMinutes,SYM,color='black',linewidth=.75)
# Axes
x_unique, x_indices = np.unique(Time[start:end,2],return_index=True)
......@@ -89,62 +101,40 @@ def Selection(Time,Date,HMC,Index_thres1,StormIndices,IndexMin1,IndexMin2,Kp,KpH
plt.show()
####################################################################
# SYM = np.loadtxt(Path_Input+'SYM_ASY_09_1981.txt',skiprows=15,usecols=[3,4,5,6])
# SYM_Time = np.linspace(HTime['dyear'][start],HTime['dyear'][end],len(SYM[30:len(SYM)-29,3]))
# # Dst
# Dst_color='dimgray'
# ax[0].plot(HTime['dyear'][start:end],Dst[start:end],color=Dst_color,linewidth=0.75,label='Dst',zorder=1)
# ax[1].plot(HTime['dyear'][start:end],Dst[start:end],color=Dst_color,zorder=1)
# ax[2].plot(HTime['dyear'][start:end],Dst[start:end],color=Dst_color,zorder=2)
# # Dst selection
# ax[0].plot(HTime['dyear'][start:end],np.ones(len(HTime['dyear'][start:end]))-30,linestyle='--',color=Dst_color,label=r'$Hl = $-30 nT',zorder=3)
# ax[0].scatter(HTime['dyear'][np.asarray(Dst_StormIndices)],Dst[np.asarray(Dst_StormIndices)],color=Dst_color,s=10,zorder=5)
# ax[1].scatter(HTime['dyear'][np.asarray(Dst_IndexMin1)],Dst[np.asarray(Dst_IndexMin1)],color=Dst_color,s=dot_size,zorder=3)
# ax[2].scatter(HTime['dyear'][np.asarray(Dst_IndexMin2)],Dst[np.asarray(Dst_IndexMin2)],color=Dst_color,s=dot_size,zorder=4)
# # Dst storm scale
# G1Dst= Rectangle((HTime['dyear'][start],-30),len(HTime['dyear'][start:end]),-20,edgecolor='yellow',facecolor='yellow',zorder=0,alpha=0.3,label='weak')
# G2Dst = Rectangle((HTime['dyear'][start],-50),len(HTime['dyear'][start:end]),-50,edgecolor='gold',facecolor='gold',zorder=0,alpha=0.3,label='moderate')
# #G3Dst = Rectangle((HTime['dyear'][start],-100),len(HTime['dyear'][start:end]),-100,edgecolor='orange',facecolor='orange',zorder=0,alpha=0.3,label='strong')
# #G4Dst = Rectangle((HTime['dyear'][start],-200),len(HTime['dyear'][start:end]),-150,edgecolor='red',facecolor='red',zorder=0,alpha=0.3,label='severe')
# #G5Dst = Rectangle((HTime['dyear'][start],-300),len(HTime['dyear'][start:end]),-600,edgecolor='maroon',facecolor='maroon',zorder=0,alpha=0.3,label='great')
# ax[2].add_patch(G1Dst);ax[2].add_patch(G2Dst)#;ax[2].add_patch(G3Dst);ax[2].add_patch(G4Dst);ax[2].add_patch(G5Dst);
# # Sym-H (only for this specific time)
# ax[3].plot(SYM_Time,SYM[30:len(SYM)-29,3],color='black',alpha=0.75)
# #ax[3].plot(SYM_Time,-SYM[30:len(SYM)-29,0],color='blue')
###
###
def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Save):
def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Dst,mode,Save):
mode = 'Kp'
TimeDecYear = Time[:,4]
# Overlap of storm times with Kp
Kp_indices = np.where(np.logical_and(KpHours_all >= min(TimeDecYear[YearsIndex]),KpHours_all < max(TimeDecYear[YearsIndex])))[0]
Kp_red = Kp_all[Kp_indices]; KpHours_red = KpHours_all[Kp_indices]
Kp_step = 0.3333333
#Kp_range = np.around(np.arange(0,9,Kp_step),3)
Kp_range = np.around(np.arange(0-Kp_step/2,9+Kp_step/2,Kp_step),3)
Storms1932 = np.where(TimeDecYear[Storms] >= 1932.000171)[0]
HMC_step = 10#7.78
Storms1957 = np.where(TimeDecYear[Storms] >= 1957.000171)[0]
HMC_step = 10
HMC_range = np.arange(0,480+HMC_step,HMC_step)
if mode == 'HMC':
if mode == 'HMC' or mode == 'Dst':
Index_range = HMC_range
step = HMC_step
else:
Index_range = Kp_range
step = Kp_step
years = np.arange(1932,2016,1)
Percentages = np.zeros((len(years),len(Index_range)))
years = np.arange(1930,2016,1)
Hist = np.zeros((len(years),len(Index_range)-1))
RHist = np.zeros((len(years),len(Index_range)-1))
KpValuesAll = []
ValuesAll = []
RValuesAll = []
for l in range(len(years)):
......@@ -186,33 +176,44 @@ def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Save):
#KpValues = Kp_hours[KpStorms,1]
#RKpValues = Kp_hours[RKpStorms,1]
HMCValues = -HMC[Storms[StormsBlock]]
RHMCValues = -HMC[RefStorms[RBlockHMC]]
if mode == 'HMC':
KpValuesAll.extend(HMCValues.tolist())
RValuesAll.extend(RHMCValues.tolist())
Hist[l,:], bin_edges = np.histogram(HMCValues, bins = Index_range)
RHist[l,:], bin_edges = np.histogram(RHMCValues, bins = Index_range)
Values = -HMC[Storms[StormsBlock]]
RValues = -HMC[RefStorms[RBlockHMC]]
elif mode == 'Dst':
Values = -Dst[Storms[StormsBlock]]
RValues = -Dst[RefStorms[RBlockHMC]]
else:
KpValuesAll.extend(KpValues.tolist())
RValuesAll.extend(RKpValues.tolist())
Hist[l,:], bin_edges = np.histogram(KpValues, bins = Index_range)
RHist[l,:], bin_edges = np.histogram(RKpValues, bins = Index_range)
Values = KpValues
RValues = RKpValues
# HMC / Kp
ValuesAll.extend(Values.tolist())
RValuesAll.extend(RValues.tolist())
Hist[l,:], bin_edges = np.histogram(Values, bins = Index_range)
RHist[l,:], bin_edges = np.histogram(RValues, bins = Index_range)
# HMC / Kp / Dst
if mode == 'HMC' or mode == 'Dst':
if mode == 'HMC':
SumHist = np.sum(Hist*100/len(Storms),axis=0)
SumHist1957 = np.sum(Hist[27:,:]*100/len(Storms1957),axis=0)
bound25_1957 = np.percentile(np.asarray(ValuesAll[27:]).flatten(),25)
bound75_1957 = np.percentile(np.asarray(ValuesAll[27:]).flatten(),75)
else:
SumHist = np.sum(Hist*100/len(Storms1957),axis=0)
else:
SumHist = np.sum(Hist*100/len(Storms1932),axis=0)
CumSum = np.cumsum(SumHist)
CumSum2 = np.cumsum(np.insert(SumHist[0:len(SumHist)-1],0,0))
bound25 = np.percentile(np.asarray(KpValuesAll).flatten(),25)
bound75 = np.percentile(np.asarray(KpValuesAll).flatten(),75)
bound50 = np.percentile(np.asarray(KpValuesAll).flatten(),50)
#print(bound25,bound50, bound75)
bound25 = np.percentile(np.asarray(ValuesAll).flatten(),25)
bound75 = np.percentile(np.asarray(ValuesAll).flatten(),75)
bound50 = np.percentile(np.asarray(ValuesAll).flatten(),50)
print(bound25,bound50, bound75)
# Reference HMC / Kp
SumRHist = np.sum(RHist*100/len(RefStorms),axis=0)
bound25_R = np.nanpercentile(np.asarray(RValuesAll).flatten(),25)
bound75_R = np.nanpercentile(np.asarray(RValuesAll).flatten(),75)
bound50_R = np.nanpercentile(np.asarray(RValuesAll).flatten(),50)
print(bound25_R, bound50_R, bound75_R)
fig = plt.figure()
fig.set_size_inches([4, 8], forward=True)
......@@ -261,8 +262,12 @@ def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Save):
axins.set_yticklabels(['0','5','10','15','20'])
####################################################################
# Dst-scale
G1Dst = Rectangle(((30-Index_range[0])/step,0),20/step,100,edgecolor='yellow',facecolor='yellow',zorder=0,alpha=0.3,label='weak')
G2Dst = Rectangle(((50-Index_range[0])/step,0),50/step,100,edgecolor='gold',facecolor='gold',zorder=0,alpha=0.3,label='moderate')
G3Dst = Rectangle(((100-Index_range[0])/step,0),100/step,100,edgecolor='orange',facecolor='orange',zorder=0,alpha=0.3,label='strong')
G4Dst = Rectangle(((200-Index_range[0])/step,0),150/step,100,edgecolor='red',facecolor='red',zorder=0,alpha=0.3,label='severe')
G5Dst = Rectangle(((350-Index_range[0])/step,0),600/step,100,edgecolor='maroon',facecolor='maroon',zorder=0,alpha=0.3)#,label='great')
# NOAA G-scale
G1 = Rectangle(((5-1.5*step-Index_range[0])/step-.5,0),3,100,edgecolor='yellow',facecolor='yellow',zorder=0,alpha=0.3,label='G1')
......@@ -271,18 +276,23 @@ def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Save):
G4 = Rectangle(((8-1.5*step-Index_range[0])/step-.5,0),4,100,edgecolor='red',facecolor='red',zorder=0,alpha=0.3,label='G4')
G5 = Rectangle(((9-0.5*step-Index_range[0])/step-.5,0),1,100,edgecolor='maroon',facecolor='maroon',zorder=0,alpha=0.3,label='G5')
if mode == 'HMC':
if mode == 'HMC' or mode == 'Dst':
ax2.bar(np.arange(0,len(Index_range)-1,1),SumHist,width=1.0,align='edge',color='darkgray',label='Target')
ax2.bar(np.arange(0,len(Index_range)-1,1),SumRHist,width=1.0,align='edge',edgecolor='midnightblue',fill=False,label='Reference')
#ax2.bar(np.arange(0,len(Index_range)-1,1),SumHist2,width=1.0,align='edge',edgecolor='maroon',fill=False,label='Target 1957')
ax2.axvline((bound25-Index_range[0])/step,color='black',linewidth=.75,linestyle='--'); ax2.axvline((bound75-Index_range[0])/step,color='black',linewidth=.75,linestyle='--',label='Q1, Q3')
ax2.axvline((bound25_R-Index_range[0])/step,color='midnightblue',linewidth=.75,linestyle='--'); ax2.axvline((bound75_R-Index_range[0])/step,color='midnightblue',linewidth=.75,linestyle='--')
#ax2.axvline((bound25_2-Index_range[0])/step,color='maroon',linestyle='--'); ax2.axvline((bound75_2-Index_range[0])/step,color='maroon',linestyle='--')
if mode == 'HMC':
ax2.bar(np.arange(0,len(Index_range)-1,1),SumHist1957,width=1.0,align='edge',edgecolor='maroon',fill=False,label='Target 1957')
ax2.axvline((bound25_1957-Index_range[0])/step,color='maroon',linewidth=.75,linestyle='--'); ax2.axvline((bound75_1957-Index_range[0])/step,color='maroon',linewidth=.75,linestyle='--')
ax2.set_xlabel('-HMC [nT]')
else:
ax2.add_patch(G1Dst);ax2.add_patch(G2Dst);ax2.add_patch(G3Dst);ax2.add_patch(G4Dst);ax2.add_patch(G5Dst)
ax2.set_xlabel('-Dst [nT]')
ax2.set_xlim([0,xlimit])
ax2.set_xticklabels(Index_range[::3], fontsize=9)
ax2.set_xlabel('-HMC [nT]')
ax2.legend(loc=1,ncol=1,frameon=False,fontsize=9)
else:
......@@ -304,8 +314,8 @@ def IndexDist(Time,YearsIndex,Storms,RefStorms,Kp_all,KpHours_all,HMC,Save):
ax2.set_ylabel('Occurrence [%]')
if Save == True:
fig.savefig('./Dump/Fig/development/KpDist_test.pdf',format='pdf',dpi=200,transparent=True)
fig.savefig('./Dump/Fig/development/KpDist_test.png',format='png',dpi=200,transparent=True)
fig.savefig('./Dump/Fig/development/'+mode+'Dist.pdf',format='pdf',dpi=200,transparent=True)
fig.savefig('./Dump/Fig/development/'+mode+'Dist.png',format='png',dpi=200,transparent=True)
plt.show()
###
###
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment