-
Notifications
You must be signed in to change notification settings - Fork 1
/
lootplotlib.py
223 lines (216 loc) · 12.6 KB
/
lootplotlib.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
@author: fscript
"""
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
import re
import numpy as np
import datetime as dt
def top10_pie(sorted_names,sorted_values,title,high_to_low=False):
if high_to_low:
idx1=0
idx2=10
step=1
else:
idx1=len(sorted_values)
idx2=len(sorted_values)-11
step=-1
theta=sorted_values[idx1:idx2:step]
beschriftung=sorted_names[idx1:idx2:step]
farbe=[(0.7-(2*ii/np.sum(sorted_values)),0.7-(2*ii/np.sum(sorted_values)),0.8) for ii in theta]
theta=np.append(theta,np.sum(sorted_values)-np.sum(sorted_values[idx1:idx2:step]))
farbe.append((0.1,0.1,0.6))
beschriftung=np.append(beschriftung,"Rest")
fig=plt.figure(figsize=(9,9))
ax=fig.add_subplot(1,1,1,aspect=1)
ax.pie(theta, labels=beschriftung,radius=1,labeldistance=1.06,colors=farbe, autopct="%.1f%%",pctdistance=0.7)
fig.suptitle(title, weight="bold")
fig.savefig(f"{re.sub('[/ ]','_',title)}.png",dpi=300)
fig.tight_layout()
plt.close(fig)
def histograms(times,values=None,var_name="Beiträge",norm_values=None):
jahresdaten=range(dt.datetime.fromtimestamp(times[0]).year,dt.datetime.fromtimestamp(times[-1]).year+1)
jahrezahlen=[f"{int(ii)}" for ii in jahresdaten]
month_bins=np.concatenate([[dt.datetime.timestamp(dt.datetime(ii,jj,1)) for jj in range(1,13)] for ii in jahresdaten])
month_bins=np.append(month_bins,dt.datetime.timestamp(dt.datetime(jahresdaten[-1]+1,1,1))-1)
jahresdaten=[dt.datetime.timestamp(dt.datetime(ii,1,1)) for ii in jahresdaten]
if type(norm_values)==type(None):
vals=values
else:
vals=np.concatenate([(np.divide(
values[(times>=month_bins[jj]) & (times<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0))],
np.sum(norm_values[(times>=month_bins[jj]) & (times<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0))])
) if np.sum(norm_values[(times>=month_bins[jj]) & (times<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0))])>0
else np.divide(np.ones(np.sum((times>=month_bins[jj]) & (times<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0)))),
-np.sum((times>=month_bins[jj]) & (times<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0)))))
for jj in range(len(month_bins)-1)])
fig = plt.figure(figsize=(9,9))
gs = plt.GridSpec(2, 1)
ax = fig.add_subplot(gs[0])
jahreszeiten = ax.hist(times, bins=month_bins, rwidth=0.8, color=(.2,.2,.8), weights=vals)#, (0,7*24*60*60),log=True)
ax.set_title(f"{var_name} pro Monat", fontdict={"fontweight":"bold"})
ax.set_ylabel(var_name)
ax.set_xticks(jahresdaten)
ax.set_xticklabels(jahrezahlen)
ax.vlines(jahresdaten,0,np.max(jahreszeiten[0])*1.05,linestyles='dashed',lw=0.8,color="k")
ax.set_ylim(0,np.max(jahreszeiten[0])*1.05)
Uhrzeiten=np.array([dt.datetime.fromtimestamp(ii).hour+(dt.datetime.fromtimestamp(ii).minute+dt.datetime.fromtimestamp(ii).second/60)/60 for ii in times])
if type(norm_values)==type(None):
vals=values
else:
month_bins=np.array([ii/2 for ii in range(49)])
# print(np.concatenate([
# np.divide(values[np.logical_and(Uhrzeiten>=month_bins[jj],Uhrzeiten<month_bins[jj+1]+(1/(24*60*60) if jj==len(month_bins)-1 else 0))],
# np.sum(norm_values[np.logical_and(Uhrzeiten>=month_bins[jj],Uhrzeiten<month_bins[jj+1]+(1/(24*60*60) if jj==len(month_bins)-1 else 0))]))
# for jj in range(5,8)]))
vals=np.concatenate([(np.divide(
values[np.logical_and(Uhrzeiten>=month_bins[jj],Uhrzeiten<month_bins[jj+1]+(1/(24*60*60) if jj==len(month_bins)-1 else 0))],
np.sum(norm_values[np.logical_and(Uhrzeiten>=month_bins[jj],Uhrzeiten<month_bins[jj+1]+(1/(24*60*60) if jj==len(month_bins)-1 else 0))])
) if np.sum(norm_values[(Uhrzeiten>=month_bins[jj]) & (Uhrzeiten<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0))])>0
else np.divide(np.ones(np.sum((Uhrzeiten>=month_bins[jj]) & (Uhrzeiten<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0)))),
-np.sum((Uhrzeiten>=month_bins[jj]) & (Uhrzeiten<month_bins[jj+1]+(1 if jj==len(month_bins)-1 else 0)))))
for jj in range(len(month_bins)-1)])
Uhrzeiten.sort() # The weights are sorted.
ax = fig.add_subplot(gs[1])
tageszeiten = ax.hist(Uhrzeiten, 24*2, (0,24), rwidth=0.8, color=(.3,.3,.8), weights=vals)
ax.set_xticks([0,2,4,6,8,10,12,14,16,18,20,22,24])
ax.set_xticklabels(["00:00","02:00","04:00","06:00","08:00","10:00","12:00","14:00","16:00","18:00","20:00","22:00","24:00"])
ax.set_title(f"{var_name} nach Tageszeit", fontdict={"fontweight":"bold"})
ax.set_xlabel("Uhrzeit")
ax.set_ylabel(var_name)
plt.text(3,np.max(tageszeiten[0]),'Jeder Balken entspricht\neiner halbe Stunde.',size=8,
horizontalalignment='center',verticalalignment='top',bbox=dict(facecolor='w',alpha=.8,edgecolor='#d0d0d0',boxstyle='round'))
ax.set_ylim(0,np.max(tageszeiten[0])*1.05)
ax.vlines([0,6,12,18,24],0,np.max(tageszeiten[0])*1.05,linestyles='dashed',lw=0.8,color="k")
fig.tight_layout()
fig.savefig(f"histogramm-{re.sub(r'[/ ]','_',var_name)}.png",dpi=300)
plt.close(fig)
def top10_histogram(top10_names,post_names,post_times,values=None,var_name="Beiträge"):
jahresdaten=range(dt.datetime.fromtimestamp(post_times[0]).year,dt.datetime.fromtimestamp(post_times[-1]).year+1)
jahrezahlen=[f"{int(ii)}" for ii in jahresdaten]
quart_bins=np.concatenate([[dt.datetime.timestamp(dt.datetime(ii,jj,1)) for jj in range(1,13,3)] for ii in jahresdaten])
quart_bins=np.append(quart_bins,dt.datetime.timestamp(dt.datetime(jahresdaten[-1]+1,1,1))-1)
jahresdaten=[dt.datetime.timestamp(dt.datetime(ii,1,1)) for ii in jahresdaten]
histposter=[post_times[post_names==ii] for ii in top10_names]
fig = plt.figure(figsize=(9,9))
gs = plt.GridSpec(3, 1)
ax = fig.add_subplot(gs[0])
ax.set_title(f"Top 10 Spieler pro Quartal: {var_name}", fontdict={"fontweight":"bold"})
jahreszeiten = ax.hist(post_times, bins=quart_bins, color="k",histtype="step", weights=values)
ax.set_xlim(quart_bins[0],quart_bins[-1]+.2*(quart_bins[-1]-[quart_bins[0]]))
ax.set_ylabel(f"Alle {var_name}")
tk=ax.get_yticks()
ax.vlines(jahresdaten,0,tk[-1],linestyles='dashed',lw=0.8,color="k")
ax.set_ylim(tk[0],tk[-1])
ax.set_yticks(tk[1:])
ax.set_xticklabels([])
if type(values)!=type(None):
top10_value=[np.concatenate([np.divide(
values[post_names==ii][np.logical_and(quart_bins[jj]<=post_times[post_names==ii],quart_bins[jj+1]>post_times[post_names==ii])],
jahreszeiten[0][jj]) for jj in range(len(quart_bins)-1)]) for ii in top10_names]
else:
top10_value=[np.concatenate([np.divide(
np.ones(np.sum([np.logical_and(quart_bins[jj]<=post_times[post_names==ii],quart_bins[jj+1]>post_times[post_names==ii])])),
jahreszeiten[0][jj]) for jj in range(len(quart_bins)-1)]) for ii in top10_names]
ax = fig.add_subplot(gs[1:])
farben=[(0.5+(i*((-1)**i))/25.0,1-float(i)/10.0,((i-5)/5.0)**2) for i in range(10)]
ax.hist(histposter, bins=quart_bins, rwidth=0.8, histtype='barstacked', label=top10_names, weights=top10_value, color=farben)
ax.set_xlim(quart_bins[0],quart_bins[-1]+.2*(quart_bins[-1]-[quart_bins[0]]))
ax.set_ylabel(f"Anteil der {var_name}")
ax.set_yticks([0,.2,.4,.6,.8,1])
ax.set_yticklabels(["0%","20%","40%","60%","80%","100%"])
ax.set_xticks(jahresdaten)
ax.set_xticklabels(jahrezahlen)
ax.vlines(jahresdaten,0,1,linestyles='dashed',lw=0.8,color="k")
ax.set_ylim(0,1)
ax.legend(loc='upper right',ncol=1)
ax.text(quart_bins[0]+.02*(quart_bins[-1]-[quart_bins[0]]),.98,'Der leere Bereich\nentspricht dem Rest.',
ha="left",va="top",size=10,bbox=dict(facecolor='w',edgecolor='#d0d0d0',boxstyle='round'))
fig.tight_layout()
fig.subplots_adjust(hspace=0)
fig.savefig(f"top10-{re.sub(r'[/ ]','_',var_name)}-histogramm.png",dpi=300)
plt.close(fig)
def time_decay_plots(post_times,life_times):
jahresdaten=range(dt.datetime.fromtimestamp(post_times[0]).year,dt.datetime.fromtimestamp(post_times[-1]).year+2)
jahrezahlen=[f"{int(ii)}" for ii in jahresdaten[:-1]]
jahresdaten=[dt.datetime.timestamp(dt.datetime(ii,1,1)) for ii in jahresdaten]
njahre=len(jahrezahlen)
if njahre<3:
fig = plt.figure(figsize=(9,9/njahre))
gs = plt.GridSpec(njahre,1)
else:
fig = plt.figure(figsize=(9,3*int(.9+njahre/3)))
gs = plt.GridSpec(int(.9+njahre/3),3)
for ii in range(njahre):
ax = fig.add_subplot(gs[ii])
c,b,r=ax.hist(
life_times[np.logical_and(jahresdaten[ii]<=post_times,jahresdaten[ii+1]>post_times)],
24*7, (0,7*24*60*60),log=True, color=(.2,.2,.8))
ax.set_xticks(range(0,8*60*60*24,60*60*24))
ax.set_xticklabels(("0","1","2","3","4","5","6","7"))
ax.set_yticks((1,10,100,1000))
ax.set_yticklabels(("1","10","100","1000"))
ax.set_title(f"Zeit nach Beitrag in {jahrezahlen[ii]}", fontdict={"fontweight":"bold"})
ax.set_xlabel("Tage")
ax.set_ylabel("Beiträge")
def linf(x,t,bb): return np.multiply(bb,np.exp(np.divide(-x,t)))
val,cov=curve_fit(linf,np.subtract(b[1:],30*60),c,p0=(60*60*4,1e3))
ax.plot(b,linf(b,*val),":", color=(.6,.6,.8),lw=.8)
ax.set_ylim(0.8,1e3)
t=int(np.log(2)*val[0])
ax.text(1.5*24*60**2,600,"$t_{1/2}$ ="+" {} Min. {} Sek.".format(int(t/60),int(t-60*int(t/60))),
size=8,bbox=dict(facecolor='w',edgecolor='#d0d0d0',boxstyle='round'))
fig.tight_layout()
plt.savefig("lifetime_decayplots.png",dpi=300)
plt.close(fig)
def top10_by_year(post_names,post_times,values=None,var_name="Beiträge",norm_values=None,min_posts_filter=3):
jahresdaten=range(dt.datetime.fromtimestamp(post_times[0]).year,dt.datetime.fromtimestamp(post_times[-1]).year+2)
jahrezahlen=[f"{int(ii)}" for ii in jahresdaten[:-1]]
jahresdaten=[dt.datetime.timestamp(dt.datetime(ii,1,1)) for ii in jahresdaten]
njahre=len(jahrezahlen)
if njahre==1:
fig = plt.figure(figsize=(9,9))
gs = plt.GridSpec(1,1)
else:
fig = plt.figure(figsize=(9,int(.9+njahre/2)*4))
gs = plt.GridSpec(int(.9+njahre/2),2)
fig.suptitle(f"Jährliche Top 10\n Gemessen an: {var_name}",weight="bold",size=24)
for jj in range(njahre):
ax = fig.add_subplot(gs[jj])
names=post_names[(post_times>=jahresdaten[jj]) & (post_times<jahresdaten[jj+1])]
if type(values)==type(None):
vals=np.ones(len(names))
else:
vals=values[(post_times>=jahresdaten[jj]) & (post_times<jahresdaten[jj+1])]
if type(norm_values)==type(None):
player_sub=np.array([(ii,np.sum(vals[names==ii])/np.sum(vals)) for ii in np.unique(names)]
,dtype=[('Spieler',"U32"),('Anteil',"float32")])
player_sub.sort(order='Anteil')
theta=np.append([1-np.sum(player_sub["Anteil"][-10:])],player_sub["Anteil"][-10:])
ax.set_xticks(np.divide(range(0,int(5*(2+np.max(theta)//.05)),5),100))
ax.set_xticklabels([f"{ii}%" for ii in range(0,int(5*(2+np.max(theta)//.05)),5)], rotation=45)
ax.set_xlim(0,int(5*(1+np.max(theta)//.05))/100)
else:
nvals=norm_values[(post_times>=jahresdaten[jj]) & (post_times<jahresdaten[jj+1])]
nn,cc=np.unique(names,return_counts=True)
player_sub=np.array([(ii,np.sum(vals[names==ii])/(np.sum(nvals[names==ii]) if np.sum(nvals[names==ii])>0 else 1))
for ii in nn[cc>min_posts_filter]]
,dtype=[('Spieler',"U32"),('Anteil',"float32")])
player_sub.sort(order='Anteil')
theta=np.append([
np.sum(vals[np.logical_not(np.isin(names,player_sub["Spieler"][:-10]))])/
np.sum(nvals[np.logical_not(np.isin(names,player_sub["Spieler"][:-10]))])],
player_sub["Anteil"][-10:])
ax.barh(range(11),theta, color=(.2,.2,.8))
ax.set_ylim(-0.8,12.5)
ax.set_yticks(range(11))
ax.set_yticklabels(np.append(["Rest"],player_sub["Spieler"][-10:]), rotation=0)
ax.set_title(f"Top 10 Spieler {jahrezahlen[jj]}", size=12, fontdict={"fontweight":"bold"})
ax.text(ax.get_xlim()[-1]/2,11.5,
f"insgesamt {len(player_sub['Spieler'])} Spieler\n und {int(np.sum(vals))} {var_name}",
horizontalalignment="center",va="center",size=10)
fig.tight_layout()
fig.savefig(f"top10byYear-{re.sub(r'[/ ]','_',var_name)}.png",dpi=300)
plt.close(fig)