aboutsummaryrefslogtreecommitdiffstats
path: root/csit.infra.dash/app/cdash/comparisons/tables.py
blob: ab99f1839d662e4eb7bd2ea97df577c6f22e86c4 (plain)
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
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
# Copyright (c) 2024 Cisco and/or its affiliates.
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at:
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

"""The comparison tables.
"""

import pandas as pd

from numpy import mean, std
from copy import deepcopy

from ..utils.constants import Constants as C
from ..utils.utils import relative_change_stdev


def select_comparison_data(
        data: pd.DataFrame,
        selected: dict,
        normalize: bool=False
    ) -> pd.DataFrame:
    """Select data for a comparison table.

    :param data: Data to be filtered for the comparison table.
    :param selected: A dictionary with parameters and their values selected by
        the user.
    :param normalize: If True, the data is normalized to CPU frequency
        Constants.NORM_FREQUENCY.
    :type data: pandas.DataFrame
    :type selected: dict
    :type normalize: bool
    :returns: A data frame with selected data.
    :rtype: pandas.DataFrame
    """

    def _calculate_statistics(
            data_in: pd.DataFrame,
            ttype: str,
            drv: str,
            norm_factor: float
        ) -> pd.DataFrame:
        """Calculates mean value and standard deviation for provided data.

        :param data_in: Input data for calculations.
        :param ttype: The test type.
        :param drv: The driver.
        :param norm_factor: The data normalization factor.
        :type data_in: pandas.DataFrame
        :type ttype: str
        :type drv: str
        :type norm_factor: float
        :returns: A pandas dataframe with: test name, mean value, standard
            deviation and unit.
        :rtype: pandas.DataFrame
        """
        d_data = {
            "name": list(),
            "mean": list(),
            "stdev": list(),
            "unit": list()
        }
        for itm in data_in["test_id"].unique().tolist():
            itm_lst = itm.split(".")
            test = itm_lst[-1].rsplit("-", 1)[0]
            if "hoststack" in itm:
                test_type = f"hoststack-{ttype}"
            else:
                test_type = ttype
            df = data_in.loc[(data_in["test_id"] == itm)]
            l_df = df[C.VALUE_ITER[test_type]].to_list()
            if len(l_df) and isinstance(l_df[0], list):
                tmp_df = list()
                for l_itm in l_df:
                    tmp_df.extend(l_itm)
                l_df = tmp_df
            try:
                mean_val = mean(l_df)
                std_val = std(l_df)
            except (TypeError, ValueError):
                continue
            d_data["name"].append(f"{test.replace(f'{drv}-', '')}-{ttype}")
            d_data["mean"].append(int(mean_val * norm_factor))
            d_data["stdev"].append(int(std_val * norm_factor))
            d_data["unit"].append(df[C.UNIT[test_type]].to_list()[0])
        return pd.DataFrame(d_data)

    lst_df = list()
    for itm in selected:
        if itm["ttype"] in ("NDR", "PDR", "Latency"):
            test_type = "ndrpdr"
        elif itm["ttype"] in ("CPS", "RPS", "BPS"):
            test_type  = "hoststack"
        else:
            test_type = itm["ttype"].lower()

        dutver = itm["dutver"].split("-", 1)  # 0 -> release, 1 -> dut version
        tmp_df = pd.DataFrame(data.loc[(
            (data["passed"] == True) &
            (data["dut_type"] == itm["dut"]) &
            (data["dut_version"] == dutver[1]) &
            (data["test_type"] == test_type) &
            (data["release"] == dutver[0])
        )])

        drv = "" if itm["driver"] == "dpdk" else itm["driver"].replace("_", "-")
        core = str() if itm["dut"] == "trex" else itm["core"].lower()
        ttype = "ndrpdr" if itm["ttype"] in ("NDR", "PDR", "Latency") \
            else itm["ttype"].lower()
        tmp_df = tmp_df[
            (tmp_df.job.str.endswith(itm["tbed"])) &
            (tmp_df.test_id.str.contains(
                (
                    f"^.*[.|-]{itm['nic']}.*{itm['frmsize'].lower()}-"
                    f"{core}-{drv}.*-{ttype}$"
                ),
                regex=True
            ))
        ]
        if itm["driver"] == "dpdk":
            for drv in C.DRIVERS:
                tmp_df.drop(
                    tmp_df[tmp_df.test_id.str.contains(f"-{drv}-")].index,
                    inplace=True
                )

        # Change the data type from ndrpdr to one of ("NDR", "PDR", "Latency")
        if test_type == "ndrpdr":
            tmp_df = tmp_df.assign(test_type=itm["ttype"].lower())

        if not tmp_df.empty:
            if normalize:
                if itm["ttype"] == "Latency":
                    norm_factor = C.FREQUENCY[itm["tbed"]] / C.NORM_FREQUENCY
                else:
                    norm_factor = C.NORM_FREQUENCY / C.FREQUENCY[itm["tbed"]]
            else:
                norm_factor = 1.0
            tmp_df = _calculate_statistics(
                tmp_df,
                itm["ttype"].lower(),
                itm["driver"],
                norm_factor
            )

        lst_df.append(tmp_df)

    if len(lst_df) == 1:
        df = lst_df[0]
    elif len(lst_df) > 1:
        df = pd.concat(
            lst_df,
            ignore_index=True,
            copy=False
        )
    else:
        df = pd.DataFrame()

    return df


def comparison_table(
        data: pd.DataFrame,
        selected: dict,
        normalize: bool,
        format: str="html"
    ) -> tuple:
    """Generate a comparison table.

    :param data: Iterative data for the comparison table.
    :param selected: A dictionary with parameters and their values selected by
        the user.
    :param normalize: If True, the data is normalized to CPU frequency
        Constants.NORM_FREQUENCY.
    :param format: The output format of the table:
        - html: To be displayed on html page, the values are shown in millions
          of the unit.
        - csv: To be downloaded as a CSV file the values are stored in base
          units.
    :type data: pandas.DataFrame
    :type selected: dict
    :type normalize: bool
    :type format: str
    :returns: A tuple with the tabe title and the comparison table.
    :rtype: tuple[str, pandas.DataFrame]
    """

    def _create_selection(sel: dict) -> list:
        """Transform the complex dictionary with user selection to list
            of simple items.

        :param sel: A complex dictionary with user selection.
        :type sel: dict
        :returns: A list of simple items.
        :rtype: list
        """
        l_infra = sel["infra"].split("-")
        selection = list()
        for core in sel["core"]:
            for fsize in sel["frmsize"]:
                for ttype in sel["ttype"]:
                    selection.append({
                        "dut": sel["dut"],
                        "dutver": sel["dutver"],
                        "tbed": f"{l_infra[0]}-{l_infra[1]}",
                        "nic": l_infra[2],
                        "driver": l_infra[-1].replace("_", "-"),
                        "core": core,
                        "frmsize": fsize,
                        "ttype": ttype
                    })
        return selection

    r_sel = deepcopy(selected["reference"]["selection"])
    c_params = selected["compare"]
    r_selection = _create_selection(r_sel)

    if format == "html" and "Latency" not in r_sel["ttype"]:
        unit_factor, s_unit_factor = (1e6, "M")
    else:
        unit_factor, s_unit_factor = (1, str())

    # Create Table title and titles of columns with data
    params = list(r_sel)
    params.remove(c_params["parameter"])
    lst_title = list()
    for param in params:
        value = r_sel[param]
        if isinstance(value, list):
            lst_title.append("|".join(value))
        else:
            lst_title.append(value)
    title = "Comparison for: " + "-".join(lst_title)
    r_name = r_sel[c_params["parameter"]]
    if isinstance(r_name, list):
        r_name = "|".join(r_name)
    c_name = c_params["value"]

    # Select reference data
    r_data = select_comparison_data(data, r_selection, normalize)

    # Select compare data
    c_sel = deepcopy(selected["reference"]["selection"])
    if c_params["parameter"] in ("core", "frmsize", "ttype"):
        c_sel[c_params["parameter"]] = [c_params["value"], ]
    else:
        c_sel[c_params["parameter"]] = c_params["value"]

    c_selection = _create_selection(c_sel)
    c_data = select_comparison_data(data, c_selection, normalize)

    if r_data.empty or c_data.empty:
        return str(), pd.DataFrame()

    l_name, l_r_mean, l_r_std, l_c_mean, l_c_std, l_rc_mean, l_rc_std, unit = \
        list(), list(), list(), list(), list(), list(), list(), set()
    for _, row in r_data.iterrows():
        if c_params["parameter"] in ("core", "frmsize", "ttype"):
            l_cmp = row["name"].split("-")
            if c_params["parameter"] == "core":
                c_row = c_data[
                    (c_data.name.str.contains(l_cmp[0])) &
                    (c_data.name.str.contains("-".join(l_cmp[2:])))
                ]
            elif c_params["parameter"] == "frmsize":
                c_row = c_data[c_data.name.str.contains("-".join(l_cmp[1:]))]
            elif c_params["parameter"] == "ttype":
                regex = r"^" + f"{'-'.join(l_cmp[:-1])}" + r"-.{3}$"
                c_row = c_data[c_data.name.str.contains(regex, regex=True)]
        else:
            c_row = c_data[c_data["name"] == row["name"]]
        if not c_row.empty:
            unit.add(f"{s_unit_factor}{row['unit']}")
            r_mean = row["mean"]
            r_std = row["stdev"]
            c_mean = c_row["mean"].values[0]
            c_std = c_row["stdev"].values[0]
            l_name.append(row["name"])
            l_r_mean.append(r_mean / unit_factor)
            l_r_std.append(r_std / unit_factor)
            l_c_mean.append(c_mean / unit_factor)
            l_c_std.append(c_std / unit_factor)
            delta, d_stdev = relative_change_stdev(r_mean, c_mean, r_std, c_std)
            l_rc_mean.append(delta)
            l_rc_std.append(d_stdev)

    s_unit = "|".join(unit)
    df_cmp = pd.DataFrame.from_dict({
        "Test Name": l_name,
        f"{r_name} Mean [{s_unit}]": l_r_mean,
        f"{r_name} Stdev [{s_unit}]": l_r_std,
        f"{c_name} Mean [{s_unit}]": l_c_mean,
        f"{c_name} Stdev [{s_unit}]": l_c_std,
        "Relative Change Mean [%]": l_rc_mean,
        "Relative Change Stdev [%]": l_rc_std
    })
    df_cmp.sort_values(
        by="Relative Change Mean [%]",
        ascending=False,
        inplace=True
    )

    return (title, df_cmp)