MTC-codes/Partitiontrace_fullL1L2.py at main · SH20025133/MTC-codes · GitHub

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
import pandas as pd
import numpy as np
import os

"""
To eliminate async writes comment out #85 and #131
"""
filename = "w87-SH-align64-sgdp7-iodepth1-clsize64-mtc.csv"

df = pd.read_csv(filename)
df.columns = ["Q1", "G1", "D1", "C1", "Q2", "G2", "D2", "C2", "src"]
#df = df[(df["I"] < df["C"]) & (df["I"] < df["D"]) & (df["D"] < df["C"])]
start = df['Q1'][0]
end = df['Q1'][len(df.index)-1]

print(start)
print(end)

df_src0 = pd.DataFrame()
df_src1 = pd.DataFrame()
df_src2 = pd.DataFrame()
df_src3 = pd.DataFrame()

start_updated = start

changepoint_BEAST = [1410]
changepoint_BEAST_mus = sorted([item*60 for item in changepoint_BEAST])

for df in pd.read_csv(filename, chunksize=1000000) :

    #df = pd.read_csv(filename)
    df.columns = ["Q1", "G1", "D1", "C1", "Q2", "G2", "D2", "C2", "src"]


    df_src0 = df_src0.append(df[(df["src"] == 0)], ignore_index=True)
    #print(df_src0.head())
    df_src1 = df_src1.append(df[(df["src"] == 1)], ignore_index=True)
    df_src2 = df_src2.append(df[(df["src"] == 2)], ignore_index=True)
    df_src3 = df_src3.append(df[(df["src"] == 3)], ignore_index=True)

    df_src0 = df_src0[(df_src0["G1"] < df_src0["C1"]) & (df_src0["G1"] < df_src0["D1"]) & (df_src0["D1"] <      df_src0["C1"])].reset_index(drop=True)
    df_src1 = df_src1[(df_src1["G1"] < df_src1["C1"]) & (df_src1["G1"] < df_src1["D1"]) & (df_src1["D1"] < df_src1["C1"]) & ((df_src1["C1"] - df_src1["D1"]) < (df_src1["G1"] < df_src1["C1"]))].reset_index(drop=True)
    df_src2 = df_src2[(df_src2["G2"] < df_src2["C2"]) & (df_src2["G2"] < df_src2["D2"]) & (df_src2["D2"] < df_src2["C2"]) & ((df_src2["C2"] - df_src2["D2"]) < (df_src2["G2"] < df_src2["C2"]))].reset_index(drop=True)
    df_src3 = df_src3[(df_src3["G1"] < df_src3["C1"]) & (df_src3["G1"] < df_src3["D1"]) & (df_src3["D1"] < df_src3["C1"]) & ((df_src3["C1"] - df_src3["D1"]) < (df_src3["G1"] < df_src3["C1"]))].reset_index(drop=True)


start0 = df_src0['Q1'][0]
start1 = df_src1['Q1'][0]
start2 = df_src2['Q2'][0]
start3 = df_src3['Q1'][0]

start_updated0 = start0
start_updated1 = start1
start_updated2 = start2
start_updated3 = start3


for item1 in changepoint_BEAST_mus:

        j = changepoint_BEAST_mus.index(item1)
        df1_src0 = df_src0[(df_src0['G1'] >= start_updated0) & (df_src0['G1'] <= start0 + item1)]
        df1_src1 = df_src1[(df_src1['G1'] >= start_updated1) & (df_src1['G1'] <= start1 + item1)]
        df1_src2 = df_src2[(df_src2['G2'] >= start_updated2) & (df_src2['G2'] <= start2 + item1)]
        df1_src3 = df_src3[(df_src3['G1'] >= start_updated3) & (df_src3['G1'] <= start3 + item1)]
        #print(df1.head())

        #df = df[(df["I"] < df["C"]) & (df["I"] < df["D"]) & (df["D"] < df["C"])]
        start_updated0 = start0 + item1
        start_updated1 = start1 + item1
        start_updated2 = start2 + item1
        start_updated3 = start3 + item1

        print("***Printing for subtrace : ", j)

        #Computing mean response time for writes (src = 0)

        try :

            Df_write_L1late = df1_src0[df1_src0["C1"] > df1_src0["C2"]]
            df_write =(Df_write_L1late.C1 - Df_write_L1late.Q1)
            Df_write_L1early = df1_src0[df1_src0["C1"] < df1_src0["C2"]]
            df_write = df_write.append(Df_write_L1early.C2 - Df_write_L1early.Q1)
            #df_write = df_write.append(df1_src1.C1 - df1_src1.Q1)
            df_write = pd.DataFrame(df_write, columns = ["RT"])
            df_write = df_write[df_write.RT < np.percentile(df_write.RT,100)]
            print("The mean response time for writes", df_write.mean()*10**6)

            #Computing mean response time for reads (src = 1,2,3)

            #df_read = (df1_src3.C1 - df1_src3.G1)*10**6
            df_read_miss = (df1_src2.C2 - df1_src2.G2)
            df_read_hit = (df1_src3.C1 - df1_src3.G1)

            df_read_hit = pd.DataFrame(df_read_hit, columns = ["RT"])
            df_read_miss = pd.DataFrame(df_read_miss, columns = ["RT"])

            df_read_hit = df_read_hit[df_read_hit.RT < np.percentile(df_read_hit.RT,100)]
            df_read_miss = df_read_miss[df_read_miss.RT < np.percentile(df_read_hit.RT,100)]

            print("The mean response time for read hits", df_read_hit.mean()*10**6)
            print("The mean response time for read misses", df_read_miss.mean()*10**6)

        except IndexError :
            continue
        #os.makedirs('Partition_traces_', exist_ok=True)
        #df1.to_csv('Partition_traces_/w09-SH-iodepth1-clsize64-mtc_'+ str(j) +'.csv',index=False)

try :
        df1_src0 = df_src0[(df_src0['G1'] >= start_updated) & (df_src0['G1'] <= start + item1)]
        df1_src1 = df_src1[(df_src1['G1'] >= start_updated) & (df_src1['G1'] <= start + item1)]
        df1_src2 = df_src2[(df_src2['G2'] >= start_updated) & (df_src2['G2'] <= start + item1)]
        df1_src3 = df_src3[(df_src3['G1'] >= start_updated) & (df_src3['G1'] <= start + item1)]

        print("df1_src0", df1_src0.head())
        print("df1_src1", df1_src1.head())
        print("df1_src2", df1_src2.head())
        print("df1_src3", df1_src3.head())

        print("***Printing for subtrace : ", j+1)

        #Computing mean response time for writes (src = 0)


        Df_write_L1late = df1_src0[df1_src0["C1"] > df1_src0["C2"]]
        df_write =(Df_write_L1late.C1 - Df_write_L1late.Q1)
        Df_write_L1early = df1_src0[df1_src0["C1"] < df1_src0["C2"]]
        df_write = df_write.append(Df_write_L1early.C2 - Df_write_L1early.Q1)
        #df_write = df_write.append(df1_src1.C1 - df1_src1.Q1)
        print("After src1", df_write.head(100))
        df_write = pd.DataFrame(df_write, columns = ["RT"])
        df_write = df_write[df_write.RT < np.percentile(df_write.RT,100)]
        print("The mean response time for writes", df_write.mean()*10**6)

        #Computing mean response time for reads (src = 1,2,3)

        #df_read = (df1_src3.C1 - df1_src3.G1)*10**6
        #print("After src1", df_read.tail())
        df_read_miss = (df1_src2.C2 - df1_src2.G2)
        df_read_hit = (df1_src3.C1 - df1_src3.G1)

        df_read_hit = pd.DataFrame(df_read_hit, columns = ["RT"])
        df_read_miss = pd.DataFrame(df_read_miss, columns = ["RT"])

        df_read_hit = df_read_hit[df_read_hit.RT < np.percentile(df_read_hit.RT,100)]
        df_read_miss = df_read_miss[df_read_miss.RT < np.percentile(df_read_hit.RT,100)]

        print("The mean response time for read hits", df_read_hit.mean()*10**6)
        print("The mean response time for read misses", df_read_miss.mean()*10**6)


except IndexError :
        pass

#os.makedirs('Partition_traces_', exist_ok=True)
#df2.to_csv('Partition_traces_/w09-SH-iodepth1-clsize64-mtc_'+ str(j+1) +'.csv',index=False)