cftime-rs 0.1.6

Rust implementation of cftime
Documentation 
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

import numpy as np
import cftime_rs
import cftime
import time
import matplotlib.pyplot as plt
from typing import Tuple, List
import os

ITERATIONS = [1, 10, 50, 100, 500, 1000, 5000, 10000, 50000, 100000, 500000, 1000000]
UNITS = "hours since 2000-01-01 00:00:00"
CALENDAR = "julian"


def performance_comparison_chart(
    cftime_rs_time: List[float], cftime_time: List[float], title: str, output_file: str
) -> None:
    """
    Generate a bar chart to compare the performance of cftime_rs and cftime.
    """

    # Calculate performance improvement percentage
    performance_improvement = [
        ((cftime_time - cftime_rs_time) / cftime_rs_time) * 100
        for cftime_time, cftime_rs_time in zip(cftime_times, cftime_rs_times)
    ]
    fig, ax1 = plt.subplots(figsize=(10, 6))
    x_ticks = np.linspace(min(ITERATIONS), max(ITERATIONS), len(ITERATIONS))
    width = ITERATIONS[-1] / len(ITERATIONS) / 2
    ax1.bar(
        x_ticks - width / 2,
        cftime_rs_times,
        width=width,
        label="cftime_rs",
        color="royalblue",
        alpha=0.7,
    )
    ax1.bar(
        x_ticks + width / 2,
        cftime_times,
        width=width,
        label="cftime",
        color="orange",
        alpha=0.7,
    )
    ax1.set_xlabel("Number of Iterations")
    ax1.set_ylabel("Execution time (seconds)")
    ax1.set_xticks(x_ticks, [str(int(x)) for x in ITERATIONS])
    ax1.set_title(title)
    ax1.set_xticks(x_ticks)
    ax1.grid(axis="y", linestyle="--", alpha=0.7)

    # Create the secondary y-axis for performance improvement
    ax2 = ax1.twinx()
    ax2.plot(
        x_ticks,
        performance_improvement,
        marker="o",
        color="green",
        label="Performance Improvement (%)",
        alpha=0.4,
    )
    ax2.set_ylabel("Performance Improvement (%)")

    ax1.legend(loc="upper left")
    ax2.legend(loc="upper right")

    plt.savefig(os.path.join(os.path.dirname(__file__), output_file))


def cftime_rs_benchmark(arr: np.array) -> float:
    cftime_rs_start = time.time()
    datetimes = cftime_rs.num2date(arr, UNITS, CALENDAR)
    for datetime in datetimes:
        datetime.__str__()
    _ = cftime_rs.date2num(datetimes, UNITS, CALENDAR, dtype="int")
    cftime_rs_end = time.time()
    return cftime_rs_end - cftime_rs_start


def cftime_benchmark(arr: np.array) -> float:
    cftime_start = time.time()
    datetimes = cftime.num2date(arr, UNITS, CALENDAR)
    for datetime in datetimes:
        datetime.__str__()
    _ = cftime.date2num(datetimes, UNITS, CALENDAR)
    cftime_end = time.time()
    return cftime_end - cftime_start


def get_data_with_str() -> Tuple[List[float], List[float]]:
    cftime_rs_times = []
    cftime_times = []
    print("cftime_rs_benchmark_with_str")
    for n in ITERATIONS:
        print(f"Number of Iterations with str: {n}")
        arr = np.array(range(n))

        cftime_rs_duration = cftime_rs_benchmark(arr)
        cftime_rs_times.append(cftime_rs_duration)
        cftime_duration = cftime_benchmark(arr)
        cftime_times.append(cftime_duration)
        print(f"cftime_rs : {cftime_rs_duration:.4f} seconds")
        print(f"cftime    : {cftime_duration:.4f} seconds")
    return cftime_rs_times, cftime_times


def cftime_rs_benchmark_without_str(arr: np.array) -> float:
    cftime_rs_start = time.time()
    datetimes = cftime_rs.num2date(arr, UNITS, CALENDAR)
    _ = cftime_rs.date2num(datetimes, UNITS, CALENDAR, dtype="int")
    cftime_rs_end = time.time()
    return cftime_rs_end - cftime_rs_start


def cftime_benchmark_without_str(arr: np.array) -> float:
    cftime_start = time.time()
    datetimes = cftime.num2date(arr, UNITS, CALENDAR)
    _ = cftime.date2num(datetimes, UNITS, CALENDAR)
    cftime_end = time.time()
    return cftime_end - cftime_start


def get_data_without_str() -> Tuple[List[float], List[float]]:
    cftime_rs_times = []
    cftime_times = []
    for n in ITERATIONS:
        print("cftime_rs_benchmark_without_str")
        print(f"Number of Iterations with str: {n}")
        arr = np.array(range(n))

        cftime_rs_duration = cftime_rs_benchmark_without_str(arr)
        cftime_rs_times.append(cftime_rs_duration)
        cftime_duration = cftime_benchmark_without_str(arr)
        cftime_times.append(cftime_duration)
        print(f"cftime_rs : {cftime_rs_duration:.4f} seconds")
        print(f"cftime    : {cftime_duration:.4f} seconds")
    return cftime_rs_times, cftime_times


def cftime_rs_benchmark_pydatetime_without_str(arr: np.array) -> float:
    cftime_rs_start = time.time()
    datetimes = cftime_rs.num2pydate(arr, UNITS, CALENDAR)
    _ = cftime_rs.pydate2num(datetimes, UNITS, CALENDAR, dtype="int")
    cftime_rs_end = time.time()
    return cftime_rs_end - cftime_rs_start


def cftime_benchmark_pydatetime_without_str(arr: np.array) -> float:
    cftime_start = time.time()
    datetimes = cftime.num2date(arr, UNITS, CALENDAR, only_use_python_datetimes=True)
    _ = cftime.date2num(datetimes, UNITS, CALENDAR)
    cftime_end = time.time()
    return cftime_end - cftime_start


def get_data_pydatetime_without_str() -> Tuple[List[float], List[float]]:
    cftime_rs_times = []
    cftime_times = []
    print("cftime_rs_benchmark_pydatetime_without_str")
    for n in ITERATIONS:
        print(f"Number of Iterations with str: {n}")
        arr = np.array(range(n))

        cftime_rs_duration = cftime_rs_benchmark_without_str(arr)
        cftime_rs_times.append(cftime_rs_duration)
        cftime_duration = cftime_benchmark_without_str(arr)
        cftime_times.append(cftime_duration)
        print(f"cftime_rs : {cftime_rs_duration:.4f} seconds")
        print(f"cftime    : {cftime_duration:.4f} seconds")
    return cftime_rs_times, cftime_times


if __name__ == "__main__":
    cftime_rs_times, cftime_times = get_data_with_str()
    performance_comparison_chart(
        cftime_rs_times,
        cftime_times,
        title="Performance Comparison: cftime_rs vs. cftime.\nDecoding, calling __str__ and encoding. \nLower is better",
        output_file="performance_comparison_with_str.png",
    )
    cftime_rs_times, cftime_times = get_data_without_str()
    performance_comparison_chart(
        cftime_rs_times,
        cftime_times,
        title="Performance Comparison: cftime_rs vs. cftime./Decoding and encoding. \nLower is better",
        output_file="performance_comparison_without_str.png",
    )
    cftime_rs_times, cftime_times = get_data_pydatetime_without_str()
    performance_comparison_chart(
        cftime_rs_times,
        cftime_times,
        title="Performance Comparison: cftime_rs vs. cftime.\nDecoding and encoding by using python datetime \nLower is better",
        output_file="performance_comparison_pydatetime_without_str.png",
    )