Skip to content

Figure 7: Average rank obtained per augmentation and dataset.

import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns

%matplotlib inline
%config InlineBackend.figure_format='retina'

import autorank
import pandas as pd

dat = {
    "mirage19": pd.read_parquet(
        "./campaigns/mirage19/augmentation-at-loading-no-dropout/minpkts10/campaign_summary/augment-at-loading/runsinfo_flowpic_dim_32.parquet"
    ),
    "mirage22_10": pd.read_parquet(
        "./campaigns/mirage22/augmentation-at-loading-no-dropout/minpkts10/campaign_summary/augment-at-loading/runsinfo_flowpic_dim_32.parquet"
    ),
    "mirage22_1000": pd.read_parquet(
        "./campaigns/mirage22/augmentation-at-loading-no-dropout/minpkts1000/campaign_summary/augment-at-loading/runsinfo_flowpic_dim_32.parquet"
    ),
    "utmobile19": pd.read_parquet(
        "./campaigns/utmobilenet21/augmentation-at-loading-no-dropout/minpkts10/campaign_summary/augment-at-loading/runsinfo_flowpic_dim_32.parquet"
    ),
}

def prepare_data(df):
    res = df[["hash", "aug_name", "seed", "split_index", "f1"]]
    res.loc[:, "id"] = (
        "split_index"
        + res.loc[:, "split_index"].astype(str)
        + "_seed"
        + res.loc[:, "seed"].astype(str)
    )
    res = res[["aug_name", "id", "f1"]]
    return res.sort_values(["aug_name", "id"])


def get_ranks(df):
    df1 = prepare_data(df)
    df1 = df1.pivot(columns="aug_name", index="id").reset_index(drop=True)
    df1.columns = df1.columns.get_level_values(1)
    new_df = pd.DataFrame(
        {
            "changertt": df1["changertt"].values,
            "colorjitter": df1["colorjitter"].values,
            "horizontalflip": df1["horizontalflip"].values,
            "noaug": df1["noaug"].values,
            "packetloss": df1["packetloss"].values,
            "rotate": df1["rotate"].values,
            "timeshift": df1["timeshift"].values,
        }
    )
    replacement = {
        "noaug": "No augmentation",
        "horizontalflip": "Horizontal flip",
        "rotate": "Rotate",
        "timeshift": "Time shift",
        "colorjitter": "Color jitter",
        "changertt": "Change RTT",
        "packetloss": "Packet Loss",
    }
    new_df = new_df.rename(columns=replacement).dropna()
    rankmat = new_df.rank(axis="columns", ascending=False)
    return rankmat

def prepare_ranks_data(dataset):
    res = get_ranks(dat[dataset])
    res["dataset"] = dataset
    return res


together = pd.concat(
    [
        prepare_ranks_data("mirage19"),
        prepare_ranks_data("mirage22_10"),
        prepare_ranks_data("mirage22_1000"),
        prepare_ranks_data("utmobile19"),
    ]
)

df_tmp = pd.melt(
    together, id_vars=["dataset"], var_name="augmentation", value_name="rank"
)

/tmp/ipykernel_13347/159222887.py:3: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
res.loc[:, "id"] = (
/tmp/ipykernel_13347/159222887.py:3: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
res.loc[:, "id"] = (
/tmp/ipykernel_13347/159222887.py:3: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
res.loc[:, "id"] = (
/tmp/ipykernel_13347/159222887.py:3: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
res.loc[:, "id"] = (

df_tmp2
dataset Mirage19 Mirage22\n(10ptks) Mirage22\n(1,000pkts) UTMobilenet21
augmentation
Change RTT 1.000000 1.733333 1.666667 3.466667
Time shift 2.733333 2.666667 3.666667 2.133333
No augmentation 3.000000 4.533333 4.933333 4.933333
Horizontal flip 3.333333 3.266667 5.200000 5.266667
Packet Loss 5.333333 3.533333 3.333333 5.933333
Color jitter 5.600000 5.533333 5.800000 3.800000
Rotate 7.000000 6.733333 3.400000 2.466667 
df_tmp2 = df_tmp.groupby(["dataset", "augmentation"])["rank"].mean().reset_index()
df_tmp2 = df_tmp2.pivot(index="augmentation", columns="dataset", values="rank")
df_tmp2 = df_tmp2.rename(
    {
        "mirage19": "Mirage19",
        "mirage22_10": "Mirage22\n(10ptks)",
        "mirage22_1000": "Mirage22\n(1,000pkts)",
        "utmobile19": "UTMobilenet21",
    },
    axis=1,
)
df_tmp2 = df_tmp2.sort_values(by="Mirage19", ascending=True)

fig, ax = plt.subplots(figsize=(5, 2.3))
ax = sns.heatmap(
    ax=ax,
    data=df_tmp2,
    annot=True,
    fmt=".1f",
    cmap=plt.get_cmap("magma_r"),
    cbar_kws=dict(label="Average rank"),
)

for lab in ax.get_yticklabels():
    if lab.get_text() in ("Change RTT", "Time shift"):
        lab.set_fontweight("bold")

ax.set_xticklabels(ax.get_xticklabels(), rotation=25, ha="right")
ax.tick_params(axis="x", which="major", pad=0)
ax.set_ylabel(None)
ax.set_xlabel(None)

plt.savefig("augmentations_rank_comparison_together2.png", bbox_inches="tight", dpi=300)

png