稻花香里说丰年

.vscode/settings.json

@@ -1,3 +1,3 @@
 {
-    "python.pythonPath": "C:\\Users\\wjs\\Anaconda3\\python.exe"
+    "python.pythonPath": "/Users/wjsjwr/.pyenv/versions/anaconda3-5.0.1/bin/python"
 }

dist_of_k.py

@@ -0,0 +1,45 @@
+import json
+from matplotlib import pyplot as plt
+from island.match import Match
+from island.matches import Matches
+from numpy import mean, std
+import numpy as np
+
+
+matches = Matches('wos-data-new')
+max_round = 15
+
+
+
+fig = plt.figure(figsize=(6.4, 3.6))
+ax = fig.gca()
+index = np.arange(11)
+ax.set_xlabel('k')
+ax.set_ylabel('p(k)')
+
+c = ['#00b894','#00cec9','#0984e3','#6c5ce7','#e84393','#d63031','#e17055','#fdcb6e','#2d3436','#6ab04c','#30336b','#ED4C67']
+
+for i in range(len(matches.data)):
+    k = np.zeros(11)
+    m = matches.data[i]
+    n = {}
+    for r in m.query('neighbor', 'create').raw_data:
+        if r['a'] in n:
+            n[r['a']].append(r['b'])
+        else:
+            n[r['a']] = [r['b']]
+
+        if r['b'] in n:
+            n[r['b']].append(r['a'])
+        else:
+            n[r['b']] = [r['a']]
+    for j in n.keys():
+        k[len(n[j])] += 1
+    pk = k / np.sum(k)
+    print(i)
+    ax.scatter(index, pk, color=c[i])
+
+# ax.set_title('Scores by group and gender')
+fig.tight_layout()
+plt.show()
+# plt.savefig('graph/neigh_per_round.eps')

island/match.py

@@ -95,6 +95,55 @@ class Match:
 
         return len(self.raw_data)
 
+
+    def get_tr(self, i, target, nb, sv):
+        """
+        # 计算tau_p的方式是统计周围邻居剩余的tr
+        # 本函数计算第i轮，所有有消耗tr的player，最后剩余的tr
+        # 由于是计算剩余tr，所以不能计算和target之间的交互
+        # nb是target的邻居
+        """
+        trs = {}
+        req = self.query('action', 'request').where(lambda x: x['rno'] == i+1)
+        app = self.query('action', 'approve').where(lambda x: x['rno'] == i+1).raw_data
+        can = self.query('action', 'cancel').where(lambda x: x['rno'] == i+1).raw_data
+        den = self.query('action', 'deny').where(lambda x: x['rno'] == i+1).raw_data
+
+        fr = []
+        for r in self.query('action', 'done').where(lambda x: x['rno'] == i and (x['a'] == target or x['b'] == target)).raw_data:
+            if r['a'] == target:
+                fr.append(r['b'])
+            elif r['b'] == target:
+                fr.append(r['a']) 
+
+        truenb = []
+        for r in nb:
+            if r not in fr and r in sv:
+                truenb.append(r)
+                trs[r] = 1440
+
+        for r in req.raw_data:
+            if r['from'] in truenb and r['to'] != target :
+                if r['from'] not in trs:
+                    trs[r['from']] = 1440
+                trs[r['from']] -= r['tr']
+
+        for r in app:
+            if r['from'] in truenb and r['to'] != target:
+                if r['from'] not in trs:
+                    trs[r['from']] = 1440
+                trs[r['from']] -= r['tr']
+
+        for r in can:
+            if r['from'] in trs and r['to'] != target:
+                trs[r['from']] += req.where(lambda x: x['from'] == r['from'] and x['to'] == r['to'] and x['log_id'] < r['log_id']).orderby('log_id').raw_data[-1]['tr']
+
+        for r in den:
+            if r['to'] in trs and r['from'] != target:
+                trs[r['to']] += req.where(lambda x: x['from'] == r['to'] and x['to'] == r['from'] and x['log_id'] < r['log_id']).orderby('log_id').raw_data[-1]['tr']
+
+        return trs
+
     @staticmethod
     def read_from_json(json_path):
         """

tau_p_co.py

@@ -10,197 +10,150 @@ from matplotlib import markers
 def error(f,x,y):
     return sp.sum((f(x)-y)**2)
 
-def get_tr(m, i, target, nb, sv):
+if __name__ == '__main__':
-    """
-    # 计算tau_p的方式是统计周围邻居剩余的tr
-    # 本函数计算第i轮，所有有消耗tr的player，最后剩余的tr
-    # 由于是计算剩余tr，所以不能计算和target之间的交互
-    # nb是target的邻居
-    """
-    trs = {}
-    req = m.query('action', 'request').where(lambda x: x['rno'] == i+1)
-    app = m.query('action', 'approve').where(lambda x: x['rno'] == i+1).raw_data
-    can = m.query('action', 'cancel').where(lambda x: x['rno'] == i+1).raw_data
-    den = m.query('action', 'deny').where(lambda x: x['rno'] == i+1).raw_data
 
-    fr = []
+    matches = Matches('wos-data-new')
-    for r in m.query('action', 'done').where(lambda x: x['rno'] == i and (x['a'] == target or x['b'] == target)).raw_data:
+    max_round = 15
-        if r['a'] == target:
-            fr.append(r['b'])
-        elif r['b'] == target:
-            fr.append(r['a']) 
 
-    truenb = []
+    survivals = {}
-    for r in nb:
+    with open('survivals.json', 'r') as f:
-        if r not in fr and r in sv:
+        survivals = json.load(f)
-            truenb.append(r)
-            trs[r] = 1440
 
-    for r in req.raw_data:
+    neighbors = {}
-        if r['from'] in truenb and r['to'] != target :
+    coopr = []
-            if r['from'] not in trs:
+    x = np.arange(1, max_round)
-                trs[r['from']] = 1440
+    mwCo = {} # Match-wise frequency of cooperation
-            trs[r['from']] -= r['tr']
+    mwTau = {} # Match-wise Tau
+    bx = []
+    tau = []
 
-    for r in app:
+    for i in range(len(matches.data)):
-        if r['from'] in truenb and r['to'] != target:
+        m = matches.data[i]
-            if r['from'] not in trs:
+        n = {}
-                trs[r['from']] = 1440
+        for r in m.query('neighbor', 'create').raw_data:
-            trs[r['from']] -= r['tr']
+            if r['a'] in n:
+                n[r['a']].append(r['b'])
+            else:
+                n[r['a']] = [r['b']]
 
-    for r in can:
+            if r['b'] in n:
-        if r['from'] in trs and r['to'] != target:
+                n[r['b']].append(r['a'])
-            trs[r['from']] += req.where(lambda x: x['from'] == r['from'] and x['to'] == r['to'] and x['log_id'] < r['log_id']).orderby('log_id').raw_data[-1]['tr']
+            else:
+                n[r['b']] = [r['a']]    
+        neighbors[matches.names[i]] = n
 
-    for r in den:
+    for i in range(max_round):
-        if r['to'] in trs and r['from'] != target:
+        co = []
-            trs[r['to']] += req.where(lambda x: x['from'] == r['to'] and x['to'] == r['from'] and x['log_id'] < r['log_id']).orderby('log_id').raw_data[-1]['tr']
+        for j in range(len(matches.data)):
+            coop = 0
+            rows = matches.data[j].query('action', 'done').where(lambda x: x['rno']==i+1).raw_data
+            for row in rows:
+                if row['act_a'] == 'C':
+                    coop += 1
+                if row['act_b'] == 'C':
+                    coop += 1
 
-    return trs
+            if rows:
+                co.append(float(coop) / float(len(rows)*2))
+                mwCo["%s-%d"%(j,i)] = co[-1]
+        bx.append(co)
+        if co:
+            coopr.append(np.average(co))
 
 
-matches = Matches('wos-data-new')
+    for i in range(max_round-1):
-max_round = 15
+        tp = []
+        for j in range(len(matches.data)):
+            if i == 0:
+                for r in matches.data[j].query('player', 'join').raw_data:
+                    t = 0
+                    k = r['pid']
+                    if k not in neighbors[matches.names[j]]:
+                        print("[%s(%d)] alone: %d" % (matches.names[j], i+1, k))
+                    else:
+                        t = 1440 * len(neighbors[matches.names[j]][k])
+                    tp.append(t if t < 1440 else 1440)
+                    mwTau["%s-%d"%(j,i)] = tp[-1]
+            else:
+                if str(i) not in survivals[matches.names[j]]:
+                    continue
+                for k in survivals[matches.names[j]][str(i)]:
+                    t = 0
+                    if k not in neighbors[matches.names[j]]:
+                        print("[%s(%d)] alone: %d" % (matches.names[j], i+1, k))
+                    else:
+                        trs = matches.data[j].get_tr(i, k, neighbors[matches.names[j]][k], survivals[matches.names[j]][str(i)])
+                        for l in neighbors[matches.names[j]][k]:
+                            if l in trs:
+                                t += trs[l]
+                    tp.append(t if t < 1440 else 1440)
+                    mwTau["%s-%d"%(j,i)] = tp[-1]
 
-survivals = {}
+        if tp:
-with open('survivals.json', 'r') as f:
+            tau.append(np.average(tp))
-    survivals = json.load(f)
-
-neighbors = {}
-coopr = []
-x = np.arange(1, max_round)
-mwCo = {} # Match-wise frequency of cooperation
-mwTau = {} # Match-wise Tau
-bx = []
-tau = []
-
-for i in range(len(matches.data)):
-    m = matches.data[i]
-    n = {}
-    for r in m.query('neighbor', 'create').raw_data:
-        if r['a'] in n:
-            n[r['a']].append(r['b'])
         else:
-            n[r['a']] = [r['b']]
+            tau.append(0)
 
-        if r['b'] in n:
+
-            n[r['b']].append(r['a'])
+    blue = '#0984e3'
+    red = '#d63031'
+
+    # p1折线图
+    # fig = plt.figure(figsize=(6.4, 3.6))
+    # ax = fig.gca()
+    # ax.plot(x, coopr, color=blue, linewidth=3)
+    # ax.set_ylim(0.5, 1)
+    # ax2 = ax.twinx()
+    # ax2.plot(x, tau, color=red, linewidth=3)
+    # ax2.set_ylim(0,1440)
+    # ax.set_xlim(1,14)
+    # ax.set_xlabel("Rounds")
+    # ax.set_ylabel("Frequency of Cooperation", color=blue)
+    # ax.tick_params(axis='y', labelcolor=blue)
+    # ax2.set_ylabel("$\\tau_{p}$", family='sans-serif', color=red)
+    # ax2.tick_params(axis='y', labelcolor=red)
+
+    # plt.tight_layout()
+    # # plt.show()
+    # plt.savefig('graph/tau_p_co_plot.eps')
+
+    tau2 = []
+    coopr2 = []
+    tau_r = []
+    coopr_r = []
+    for i in range(len(tau)):
+        if tau[i] <= 720:
+            tau2.append(tau[i])
+            coopr2.append(coopr[i])
         else:
-            n[r['b']] = [r['a']]    
+            tau_r.append(tau[i])
-    neighbors[matches.names[i]] = n
+            coopr_r.append(coopr[i])
 
-for i in range(max_round):
+    # p2散点图
-    co = []
+    fig = plt.figure(figsize=(6.4, 3.6))
-    for j in range(len(matches.data)):
+    ax = fig.gca()
-        coop = 0
+    # ax.set_ylim(0.5, 1)
-        rows = matches.data[j].query('action', 'done').where(lambda x: x['rno']==i+1).raw_data
+    fp1,residuals,rank,sv,rcond = sp.polyfit(tau2, coopr2, 1, full=True)
-        for row in rows:
+    print("残差：",residuals)
-            if row['act_a'] == 'C':
+    print('Model parameter:',fp1)
-                coop += 1
+    f1 = sp.poly1d(fp1)
-            if row['act_b'] == 'C':
+    print("error= %f" % error(f1, tau2, coopr2))
-                coop += 1
+    # fx = sp.linspace(0,max(tau2),1000)
+    fx = sp.linspace(0,720,2)
 
-        if rows:
+    plt.plot(fx,f1(fx),linewidth=2,color=red, ls='--', zorder=0)
-            co.append(float(coop) / float(len(rows)*2))
+    plt.scatter(tau2, coopr2, color=blue, linewidths=2, zorder=100)
-            mwCo["%s-%d"%(j,i)] = co[-1]
+    plt.scatter(tau_r, coopr_r, color='white', edgecolors=blue, linewidths=2, zorder=101)
-    bx.append(co)
+    ax.set_xlabel('$\\tau_{p}$', family='sans-serif')
-    if co:
+    ax.set_ylabel('Frequency of Cooperation')
-        coopr.append(np.average(co))
+    ax.set_xlim(0, 1440)
-
+    ax.set_xticks(sp.linspace(0, 1440, 13))
-
+    ax.set_ylim(0.5, 1)
-for i in range(max_round-1):
+    plt.tight_layout()
-    tp = []
+    plt.show()
-    for j in range(len(matches.data)):
+    # plt.savefig('graph/tau_p_co_sca.eps')
-        if i == 0:
-            for r in matches.data[j].query('player', 'join').raw_data:
-                t = 0
-                k = r['pid']
-                if k not in neighbors[matches.names[j]]:
-                    print("[%s(%d)] alone: %d" % (matches.names[j], i+1, k))
-                else:
-                    t = 1440 * len(neighbors[matches.names[j]][k])
-                tp.append(t if t < 1440 else 1440)
-                mwTau["%s-%d"%(j,i)] = tp[-1]
-        else:
-            if str(i) not in survivals[matches.names[j]]:
-                continue
-            for k in survivals[matches.names[j]][str(i)]:
-                t = 0
-                if k not in neighbors[matches.names[j]]:
-                    print("[%s(%d)] alone: %d" % (matches.names[j], i+1, k))
-                else:
-                    trs = get_tr(matches.data[j], i, k, neighbors[matches.names[j]][k], survivals[matches.names[j]][str(i)])
-                    for l in neighbors[matches.names[j]][k]:
-                        if l in trs:
-                            t += trs[l]
-                tp.append(t if t < 1440 else 1440)
-                mwTau["%s-%d"%(j,i)] = tp[-1]
-
-    if tp:
-        tau.append(np.average(tp))
-    else:
-        tau.append(0)
-
-
-blue = '#0984e3'
-red = '#d63031'
-
-# p1折线图
-# fig = plt.figure(figsize=(6.4, 3.6))
-# ax = fig.gca()
-# ax.plot(x, coopr, color=blue, linewidth=3)
-# ax.set_ylim(0.5, 1)
-# ax2 = ax.twinx()
-# ax2.plot(x, tau, color=red, linewidth=3)
-# ax2.set_ylim(0,1440)
-# ax.set_xlim(1,14)
-# ax.set_xlabel("Rounds")
-# ax.set_ylabel("Frequency of Cooperation", color=blue)
-# ax.tick_params(axis='y', labelcolor=blue)
-# ax2.set_ylabel("$\\tau_{p}$", family='sans-serif', color=red)
-# ax2.tick_params(axis='y', labelcolor=red)
-
-# plt.tight_layout()
-# # plt.show()
-# plt.savefig('graph/tau_p_co_plot.eps')
-
-tau2 = []
-coopr2 = []
-tau_r = []
-coopr_r = []
-for i in range(len(tau)):
-    if tau[i] <= 720:
-        tau2.append(tau[i])
-        coopr2.append(coopr[i])
-    else:
-        tau_r.append(tau[i])
-        coopr_r.append(coopr[i])
-
-# p2散点图
-fig = plt.figure(figsize=(6.4, 3.6))
-ax = fig.gca()
-# ax.set_ylim(0.5, 1)
-fp1,residuals,rank,sv,rcond = sp.polyfit(tau2, coopr2, 1, full=True)
-print("残差：",residuals)
-print('Model parameter:',fp1)
-f1 = sp.poly1d(fp1)
-print("error= %f" % error(f1, tau2, coopr2))
-# fx = sp.linspace(0,max(tau2),1000)
-fx = sp.linspace(0,720,2)
-
-plt.plot(fx,f1(fx),linewidth=2,color=red, ls='--', zorder=0)
-plt.scatter(tau2, coopr2, color=blue, linewidths=2, zorder=100)
-plt.scatter(tau_r, coopr_r, color='white', edgecolors=blue, linewidths=2, zorder=101)
-ax.set_xlabel('$\\tau_{p}$', family='sans-serif')
-ax.set_ylabel('Frequency of Cooperation')
-ax.set_xlim(0, 1440)
-ax.set_xticks(sp.linspace(0, 1440, 13))
-ax.set_ylim(0.5, 1)
-plt.tight_layout()
-plt.show()
-# plt.savefig('graph/tau_p_co_sca.eps')
 
 
 
-# 皮尔逊相关系数
+    # 皮尔逊相关系数
-print("pearson: %f, p-value: %f" % pearsonr(tau2, coopr2))
+    print("pearson: %f, p-value: %f" % pearsonr(tau2, coopr2))

taup_and_new_partner.py

@@ -0,0 +1,122 @@
+import json
+from matplotlib import pyplot as plt
+from island.match import Match
+from island.matches import Matches
+import numpy as np
+from scipy.stats import pearsonr
+
+matches = Matches('wos-data-new')
+
+k = np.arange(0, 1441, 144)
+succ = np.zeros(11)
+total = np.zeros(11)
+
+survivals = {}
+with open('survivals.json', 'r') as f:
+    survivals = json.load(f)
+
+neighbors = {}
+for i in range(len(matches.data)):
+    m = matches.data[i]
+    n = {}
+    for r in m.query('neighbor', 'create').raw_data:
+        if r['a'] in n:
+            n[r['a']].append(r['b'])
+        else:
+            n[r['a']] = [r['b']]
+
+        if r['b'] in n:
+            n[r['b']].append(r['a'])
+        else:
+            n[r['b']] = [r['a']]    
+    neighbors[matches.names[i]] = n
+
+for m_i in range(len(matches.data)):
+    m = matches.data[m_i]
+    info = m.query('game', 'created').select('info').first()['info']
+    conf = json.loads(info['config'])
+    game_end_at = int(info['game_end_at'])
+
+    for p in m.query('player', 'join').raw_data:
+        pid = p['pid']
+        for i in range(2, game_end_at):
+            neighborhood = []
+            if pid not in neighbors[matches.names[m_i]]:
+                break
+            for j in neighbors[matches.names[m_i]][pid]:
+                if j in survivals[matches.names[m_i]][str(i-1)]:
+                    neighborhood.append(j)
+            if len(neighborhood) < 2:
+                break
+            previous_round_partner = []
+            for r in m.query('action', 'done').where(lambda x: x['rno']==i-1 and (x['a']==pid or x['b']==pid)).raw_data:
+                if r['a'] == pid:
+                    previous_round_partner.append(r['b'])
+                else:
+                    previous_round_partner.append(r['a'])
+            new_partner_request = 0
+            for r in m.query('action', 'request').where(lambda x: x['rno']==i and (x['from']==pid or x['to']==pid)).raw_data:
+                if r['from'] == pid:
+                    if r['to'] not in previous_round_partner:
+                        new_partner_request += 1
+                else:
+                    if r['from'] not in previous_round_partner:
+                        new_partner_request += 1
+            if new_partner_request == 0:
+                continue
+            new_partner_succ = 0
+            for r in m.query('action', 'done').where(lambda x: x['rno']==i and (x['a']==pid or x['b']==pid)).raw_data:
+                if r['a'] == pid:
+                    if r['b'] not in previous_round_partner:
+                        new_partner_succ += 1
+                else:
+                    if r['a'] not in previous_round_partner:
+                        new_partner_succ += 1
+            t = 0
+            trs = m.get_tr(i, pid, neighbors[matches.names[m_i]][pid], survivals[matches.names[m_i]][str(i-1)])
+            for l in neighborhood:
+                if l in trs:
+                    t += trs[l]
+            t = t if t <= 1440 else 1440
+            succ[t//144] += new_partner_succ
+            total[t//144] += new_partner_request
+
+red = '#d63031'
+fig = plt.figure(figsize=(6.4, 4))
+ax = fig.gca()
+bar_width = 0.35
+opacity = 1
+error_config = {'ecolor': '0.3', 'capsize': 4}
+# rects1 = ax.bar(k, succ, bar_width,
+#                 alpha=opacity, color='#00b894',
+#                 # yerr=c_req_suc_std, error_kw=error_config,
+#                 label='Success')
+# rects3 = ax.bar(k + bar_width, total, bar_width,
+#                 alpha=opacity, color='#fdcb6e',
+#                 # yerr=d_req_suc_mean, error_kw=error_config,
+#                 label='Requests')
+
+ax.set_xlabel('k')
+ax.set_ylabel('Count')
+# ax.set_title('Scores by group and gender')
+# ax.set_xticks(k + bar_width / 2)
+# ax.set_xticklabels(k)
+ax.legend()
+ax2 = ax.twinx()
+ax.plot(k, succ, color='#00b894', lw=2, marker='*', ls='none')
+ax.plot(k, total, color='#fdcb6e', lw=2, marker='+', ls='none')
+for i in range(11):
+    if total[i] == 0:
+        total[i] = 1
+ax2.plot(k, succ/total,linewidth=2,color=red, ls='--')
+# ax2.set_ylabel("Frequency of new partners", family='sans-serif', color=red)
+# ax2.tick_params(axis='y', labelcolor=red)
+ax2.set_ylim(0,1)
+fig.tight_layout()
+plt.show()
+# plt.savefig('graph/k_and_new_partner.eps')
+
+# print("[succ vs k]pearson: %f, p-value: %f" % pearsonr(succ, k))
+# print("[total vs k]pearson: %f, p-value: %f" % pearsonr(total, k))
+# print("[rate vs k]pearson: %f, p-value: %f" % pearsonr(succ/total, k))
+# print(np.average(succ/total))