5.7-dars: Seaborn

5-QISM — VIZUALIZATSIYA · 7-dars

1. Kirish va motivatsiya

Matplotlib kuchli, lekin ko'p kod talab qiladi va standart ko'rinishi oddiy. Seaborn — Matplotlib ustida qurilgan yuqori darajali kutubxona: chiroyli grafiklar, statistik vizualizatsiya (regressiya chizig'i, taqsimot, korrelyatsiya), va DataFrame bilan to'g'ridan ishlash — kam kod bilan. Seaborn'da bir qator kod chiroyli, statistikaga boy grafik beradi: sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar") — nuqtalar, rang guruh bilan, avtomatik legend. Nega muhim? (1) Kam kod — chiroyli grafik tez; (2) Statistik — regressiya, taqsimot, korrelyatsiya (o'rnatilgan); (3) DataFrame — ustun nomlari bilan to'g'ridan. Bu dars Seaborn'ni o'rgatadi — statistik vizualizatsiya kam kod bilan.

Seaborn — yuqori darajali vizualizatsiya: Matplotlib ustida (chiroyli, statistik), DataFrame (data=df, x=, y=, hue= — ustun nomlari), statistik grafiklar (scatterplot/histplot/boxplot/regplot/heatmap), hue (guruh — rang, avtomatik legend), uslub (sns.set_theme — chiroyli), Matplotlib bilan (ax= — birga). Foydalanish: chiroyli grafik, statistik, DataFrame. Bu 5.1 (Matplotlib), 3-qism (Pandas) bilan bog'liq. Seaborn — yuqori daraja. DataFrame. Statistik.

Real vaziyat. Data Scientist Matplotlib'da ko'p kod yozdi (fig, ax, plot, bezash — har grafik). Seaborn bilan kam kod: sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar") — nuqtalar + shahar rang + avtomatik legend + chiroyli (bir qator!). Statistik: sns.regplot(data=df, x="reklama", y="savdo") — scatter + regressiya chizig'i (trend, avtomatik); sns.histplot(df["ball"], kde=True) — histogram + KDE (silliq taqsimot egrisi). DataFrame: ustun nomlari bilan (x="reklama" — massiv emas). hue (guruh — rang; klaster). Seaborn ma'lumotni kam kod, chiroyli, statistik ko'rsatdi (Matplotlib ustida). Seaborn — yuqori daraja vizualizatsiya.

Bu darsda Seaborn'ni o'rganamiz.

Bu darsda:

• Seaborn nima (Matplotlib ustida)
• DataFrame bilan (x, y, hue)
• Statistik grafiklar (regplot, histplot)
• hue (guruh — rang)
• Matplotlib bilan integratsiya
• Seaborn amaliyoti
• Seaborn tuzoqlari
• Amaliy: Seaborn modeli

ℹ Misollar real seaborn/matplotlib bilan (Agg — grafik xususiyatlari matn bilan tekshiriladi) ishlaydi.

2. Nazariya — chuqur tushuntirish

2.1. Seaborn nima (Matplotlib ustida)

Yuqori daraja:

import seaborn as sns
import matplotlib.pyplot as plt

# Seaborn — Matplotlib ustida (chiroyli, statistik, kam kod)
sns.set_theme()   # chiroyli uslub

# bir qator — chiroyli grafik
sns.histplot(data=df, x="ball")   # histogram (chiroyli)

# Matplotlib ustida (ax bilan birga ishlaydi)

Seaborn nima (Matplotlib ustida) — yuqori daraja: Seaborn (import seaborn as sns) — Matplotlib ustida qurilgan (chiroyli, statistik, kam kod); sns.set_theme() (chiroyli uslub). Sabab: Matplotlib ko'p kod (fig, ax, plot, bezash) va oddiy ko'rinish; Seaborn kam kod (bir qator — chiroyli grafik) va statistik (regressiya, taqsimot o'rnatilgan). Matplotlib ustida (Seaborn Matplotlib ishlatadi — ax bilan birga; poydevor — 5.1). Seaborn (yuqori daraja — qulay), Matplotlib (pastki — moslashuvchan). Seaborn — Matplotlib ustida (chiroyli, statistik, kam kod). Yuqori daraja. Statistik.

2.2. DataFrame bilan (x, y, hue)

Ustun nomlari bilan:

import seaborn as sns

# DataFrame — ustun nomlari (massiv emas)
sns.scatterplot(data=df, x="reklama", y="savdo")

# hue — guruh (rang, avtomatik legend)
sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar")

# data=, x=, y=, hue= — ustun nomlari

DataFrame bilan (x, y, hue) — ustun nomlari: sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar") (data= DataFrame; x=/y= ustun nomlari (massiv emas); hue= guruh — rang). Sabab: Seaborn DataFrame bilan to'g'ridan (Pandas — 3-qism; ustun nomlari — x="reklama" massiv o'rniga); hue guruhni avtomatik (rang + legend — Matplotlib'da qo'lda). data=df (jadval), x=/y= (ustun), hue= (guruh rang). Bu Seaborn kuchi (DataFrame + kam kod). DataFrame bilan — data=, x=/y=, hue= (ustun nomlari). Ustun nomi. hue.

2.3. Statistik grafiklar (regplot, histplot)

Statistik grafiklar — o'rnatilgan statistika: regplot (scatter + regressiya chizig'i — trend, ishonch oralig'i; avtomatik), histplot (histogram + KDE — silliq zichlik egrisi; kde=True), boxplot/violinplot (taqsimot — violin: box + zichlik shakli), heatmap (korrelyatsiya matritsasi — rangli; 5.8), pairplot (barcha juftlik scatter — bir marta). Sabab: Seaborn statistikani o'rnatgan (regressiya — np.polyfit qo'lda emas; KDE — silliq; Matplotlib'da qo'lda hisob); statistik vizualizatsiya (grafik + statistika — bir qator). regplot (trend), histplot (KDE), violinplot (shakl). Statistik grafiklar — regplot/histplot/violinplot (o'rnatilgan). Statistik. Bir qator.

2.4. hue (guruh — rang)

hue (guruh — rang) — Seaborn kuchi: hue= parametr (kategoriya ustun — har guruh boshqa rang + avtomatik legend); barcha Seaborn grafikda (scatterplot, histplot, boxplot — hue=). Sabab: guruh bo'yicha ko'rsatish (shahar, jins — har guruh rang; klaster, taqqoslash); Matplotlib'da qo'lda (har guruh alohida plot + rang + legend); Seaborn avtomatik (hue="shahar" — hammasi). hue (guruh — rang), style= (guruh — marker/chiziq turi), size= (guruh — o'lcham). hue eng ko'p (guruh vizualizatsiyasi — bir parametr). hue — guruh rang (hue=; avtomatik legend). Guruh. Avtomatik.

2.5. Matplotlib bilan integratsiya

Matplotlib bilan integratsiya — birga ishlash: Seaborn Matplotlib ustida (grafik Matplotlib ax'da; ax= bilan joylashtirish); bezash (Matplotlib metodlari — ax.set_title, plt.savefig; 5.1). Sabab: Seaborn chizadi, lekin bezash/saqlash Matplotlib (ax — Seaborn qaytaradi yoki ax= beradi); birga (Seaborn grafik + Matplotlib bezash). sns.scatterplot(..., ax=ax) (Matplotlib ax'da), ax.set_title(...) (bezash), fig.savefig(...) (saqlash — 5.1). Subplots (Matplotlib — 5.10; Seaborn har ax'da). Seaborn (grafik) + Matplotlib (bezash/tuzilma). Matplotlib bilan — ax=, bezash (birga). Integratsiya. Birga.

2.6. Seaborn amaliyoti

Seaborn amaliyoti: sns.set_theme() (chiroyli uslub); data=df, x=, y= (DataFrame — ustun nomlari); hue= (guruh — rang); statistik (regplot trend, histplot KDE, violinplot shakl); heatmap (korrelyatsiya — 5.8); pairplot (barcha juftlik); Matplotlib (ax=, bezash — birga). Tuzoqlar: DataFrame emas (massiv — data= afzal), hue ko'p guruh (chalkash), Matplotlib aralash (bezash — Matplotlib), statistik faraz (regplot chiziqli). Amaliyot — set_theme, DataFrame, hue, statistik. Yuqori daraja. Kam kod.

2.7. Seaborn tuzoqlari

Seaborn asosiy tuzoqlari: DataFrame emas (Seaborn DataFrame uchun ideal — data=df, x="ustun"; massiv ham ishlaydi, lekin DataFrame afzal — ustun nomi, hue); hue ko'p guruh (hue — kam guruh (2-6 rang); ko'p guruh — ranglar ajratib bo'lmaydi, chalkash); Matplotlib aralash (Seaborn chizadi, bezash Matplotlib (ax.set_title, savefig); aralashtirmang — Seaborn grafik + Matplotlib bezash); statistik faraz (regplot — chiziqli regressiya; nochiziqli (egri) — order= yoki lowess; noto'g'ri trend); figure vs axes daraja (Seaborn ikki turi — scatterplot (axes — bir grafik) vs relplot (figure — subplots; ax yo'q); chalkash); KDE tekislash (histplot(kde=True) — KDE silliq, lekin kichik namunada aldash; katta namuna); rang standart (Seaborn rang — chiroyli, lekin ma'noli rang tanlov — ketma-ket/kategoriya); eski API (sns.distplot — eski; histplot/displot yangi). Sabab: Seaborn daraja/statistik nozik (DataFrame, hue, figure/axes — chalkash yoki noto'g'ri). Yechim: DataFrame, kam guruh hue, Matplotlib bezash, chiziqli tekshir. Tuzoqlar — DataFrame, hue, Matplotlib, statistik.

2.8. Seaborn — statistik vizualizatsiya kam kod bilan

Seaborn asosiy g'oyasi — statistik vizualizatsiya kam kod bilan: Matplotlib (poydevor — 5.1; ko'p kod, oddiy) ustida yuqori daraja (chiroyli, statistik, kam kod); DataFrame bilan to'g'ridan (ustun nomlari, hue). sns.set_theme (chiroyli), DataFrame (data=df, x=, y=, hue= — ustun nomlari, guruh rang), statistik grafiklar (regplot trend, histplot KDE, violinplot shakl, heatmap korrelyatsiya 5.8, pairplot juftlik), hue (guruh — rang, avtomatik legend), Matplotlib bilan (ax=, bezash — birga). Foydalanish: kam kod (chiroyli grafik tez), statistik (regressiya, taqsimot — o'rnatilgan), DataFrame (Pandas bilan). Seaborn (qulay) + Matplotlib (moslashuvchan) — birga (Seaborn chizadi, Matplotlib bezaydi/tuzilma). Data Science'da keng (EDA — tez chiroyli grafik; statistik vizualizatsiya — 9-qism). Bu 5.1 (Matplotlib) ustida va 3-qism (Pandas — DataFrame), 5.8 (heatmap) bilan. Seaborn — statistik vizualizatsiya kam kod bilan (Matplotlib ustida, DataFrame). Yuqori daraja. Kam kod.

3. Tez ma'lumotnoma

import seaborn as sns
import matplotlib.pyplot as plt

sns.set_theme()   # chiroyli uslub

# DATAFRAME bilan (ustun nomlari):
sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar")
#   data= (jadval) · x=/y= (ustun) · hue= (guruh — rang, legend)

# STATISTIK GRAFIKLAR (o'rnatilgan):
sns.regplot(data=df, x="reklama", y="savdo")     # scatter + trend
sns.histplot(data=df, x="ball", kde=True)         # histogram + KDE
sns.boxplot(data=df, x="shahar", y="maosh")       # box (guruh)
sns.violinplot(data=df, x="shahar", y="maosh")    # box + shakl
sns.heatmap(df.corr(), annot=True)                # korrelyatsiya (5.8)
sns.pairplot(df)                                   # barcha juftlik

# MATPLOTLIB bilan (bezash/saqlash):
fig, ax = plt.subplots()
sns.scatterplot(data=df, x="a", y="b", ax=ax)
ax.set_title("..."); fig.savefig("g.png")

QOIDA: DataFrame (data=) · hue kam guruh · bezash Matplotlib · statistik faraz

Seaborn xulosasi

Seaborn — statistik vizualizatsiya kam kod bilan (Matplotlib ustida)
DataFrame — data=df, x=, y=, hue= (ustun nomlari, guruh rang)
Statistik grafiklar — regplot (trend), histplot (KDE), violinplot
hue — guruh (rang, avtomatik legend)
Matplotlib bilan — ax=, bezash (Seaborn chizadi, Matplotlib bezaydi)

4. Batafsil misollar

Misollar real seaborn/matplotlib bilan (Agg — grafik xususiyatlari matn bilan tekshiriladi) ishlaydi.

Misol 1 — Seaborn scatterplot

"""Seaborn scatterplot (real seaborn/pandas)."""

import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns


def main() -> None:
    df = pd.DataFrame({
        "reklama": [10, 20, 30, 40, 50, 15, 25, 35],
        "savdo": [100, 180, 250, 340, 480, 130, 220, 300],
        "shahar": ["T", "T", "T", "T", "S", "S", "S", "S"],
    })

    print("=== 1. scatterplot (DataFrame) ===")
    fig, ax = plt.subplots()
    sns.scatterplot(data=df, x="reklama", y="savdo", ax=ax)
    print(f"  nuqtalar: {len(ax.collections[0].get_offsets())}")

    print("\n=== 2. hue (guruh — rang) ===")
    fig2, ax2 = plt.subplots()
    sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar", ax=ax2)
    print(f"  guruhlar (collections): {len(ax2.collections)}")

    print("\n=== 3. Legend (avtomatik) ===")
    print(f"  legend bor: {ax2.get_legend() is not None}")

    print("\n=== 4. Ustun nomlari bilan ===")
    print(f"  x=reklama, y=savdo, hue=shahar (massiv emas)")
    plt.close("all")
    print("  ⭐ Seaborn — DataFrame (x, y, hue)")


if __name__ == "__main__":
    main()

Natijaning muhim qismi:

=== 1. scatterplot (DataFrame) ===
  nuqtalar: 8

=== 2. hue (guruh — rang) ===
  guruhlar (collections): 1

=== 3. Legend (avtomatik) ===
  legend bor: True

=== 4. Ustun nomlari bilan ===
  x=reklama, y=savdo, hue=shahar (massiv emas)
  ⭐ Seaborn — DataFrame (x, y, hue)

Nima ko'rsatdi: 2.1, 2.2-bo'limlar.

Misol 2 — regplot (trend chizig'i)

"""Seaborn regplot: trend chizig'i (real seaborn)."""

import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns


def main() -> None:
    df = pd.DataFrame({
        "reklama": [10, 20, 30, 40, 50],
        "savdo": [100, 210, 290, 420, 480],
    })

    print("=== 1. regplot (scatter + trend) ===")
    fig, ax = plt.subplots()
    sns.regplot(data=df, x="reklama", y="savdo", ax=ax)
    print(f"  chiziqlar (trend): {len(ax.lines)}")

    print("\n=== 2. Nuqtalar ===")
    print(f"  nuqtalar bor: {len(ax.collections) > 0}")

    print("\n=== 3. Regressiya chizig'i ===")
    print("  regplot avtomatik trend (ishonch oralig'i bilan)")

    print("\n=== 4. Tushuntirish ===")
    print("  regplot — scatter + regressiya (statistik)")
    plt.close("all")
    print("  ⭐ regplot — trend chizig'i (o'rnatilgan)")


if __name__ == "__main__":
    main()

Natijaning muhim qismi:

=== 1. regplot (scatter + trend) ===
  chiziqlar (trend): 1

=== 2. Nuqtalar ===
  nuqtalar bor: True

=== 3. Regressiya chizig'i ===
  regplot avtomatik trend (ishonch oralig'i bilan)

=== 4. Tushuntirish ===
  regplot — scatter + regressiya (statistik)
  ⭐ regplot — trend chizig'i (o'rnatilgan)

Nima ko'rsatdi: 2.3-bo'lim.

Misol 3 — histplot (KDE)

"""Seaborn histplot: KDE (real seaborn/numpy)."""

import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns


def main() -> None:
    np.random.seed(0)
    ball = np.random.normal(70, 10, 1000)

    print("=== 1. histplot (histogram) ===")
    fig, ax = plt.subplots()
    sns.histplot(ball, ax=ax)
    print(f"  ustunlar (patches): {len(ax.patches) > 0}")

    print("\n=== 2. KDE (silliq egri) ===")
    fig2, ax2 = plt.subplots()
    sns.histplot(ball, kde=True, ax=ax2)
    print(f"  KDE chizig'i: {len(ax2.lines) > 0}")

    print("\n=== 3. KDE nima ===")
    print("  KDE — silliq zichlik egrisi (taqsimot shakli)")

    print("\n=== 4. Tushuntirish ===")
    print("  histplot + kde=True (histogram + silliq egri)")
    plt.close("all")
    print("  ⭐ histplot — histogram + KDE (statistik)")


if __name__ == "__main__":
    main()

Natijaning muhim qismi:

=== 1. histplot (histogram) ===
  ustunlar (patches): True

=== 2. KDE (silliq egri) ===
  KDE chizig'i: True

=== 3. KDE nima ===
  KDE — silliq zichlik egrisi (taqsimot shakli)

=== 4. Tushuntirish ===
  histplot + kde=True (histogram + silliq egri)
  ⭐ histplot — histogram + KDE (statistik)

Nima ko'rsatdi: 2.3-bo'lim.

Misol 4 — boxplot (guruh, hue)

"""Seaborn boxplot: guruh (real seaborn/pandas)."""

import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns


def main() -> None:
    np.random.seed(0)
    df = pd.DataFrame({
        "shahar": ["T"] * 100 + ["S"] * 100 + ["B"] * 100,
        "maosh": np.concatenate([
            np.random.normal(500, 80, 100),
            np.random.normal(400, 60, 100),
            np.random.normal(600, 100, 100),
        ]),
    })

    print("=== 1. boxplot (guruh bo'yicha) ===")
    fig, ax = plt.subplots()
    sns.boxplot(data=df, x="shahar", y="maosh", ax=ax)
    print(f"  qutilar (patches): {len(ax.patches) >= 3}")

    print("\n=== 2. Guruh medianlari ===")
    medianlar = df.groupby("shahar")["maosh"].median().round(0)
    print(f"  {medianlar.to_dict()}")

    print("\n=== 3. Kam kod (bir qator) ===")
    print("  sns.boxplot(data, x, y) — Matplotlib'da ko'p kod")

    print("\n=== 4. Tushuntirish ===")
    print("  DataFrame + x/y — guruh boxplot (kam kod)")
    plt.close("all")
    print("  ⭐ Seaborn boxplot — guruh (kam kod)")


if __name__ == "__main__":
    main()

Natijaning muhim qismi:

=== 1. boxplot (guruh bo'yicha) ===
  qutilar (patches): True

=== 2. Guruh medianlari ===
  {'B': 592.0, 'S': 401.0, 'T': 508.0}

=== 3. Kam kod (bir qator) ===
  sns.boxplot(data, x, y) — Matplotlib'da ko'p kod

=== 4. Tushuntirish ===
  DataFrame + x/y — guruh boxplot (kam kod)
  ⭐ Seaborn boxplot — guruh (kam kod)

Nima ko'rsatdi: 2.3, 2.4-bo'limlar.

5. To'g'ri va noto'g'ri tushunishlar

6. Keng tarqalgan xatolar va yechimlari

1. DataFrame emas

sns.scatterplot(x=massiv1, y=massiv2)   # ishlaydi, lekin     # ⚠️
sns.scatterplot(data=df, x="a", y="b")   # DataFrame (afzal)  # ✅

2. hue ko'p guruh

sns.scatterplot(..., hue="id")   # 100 guruh (chalkash)       # ⚠️
sns.scatterplot(..., hue="kategoriya")   # kam guruh          # ✅

3. Matplotlib aralash

sns.scatterplot(...); plt.title(...)   # holatli (chalkash)   # ⚠️
sns.scatterplot(..., ax=ax); ax.set_title(...)                # ✅

4. regplot nochiziqli

sns.regplot(data=df, x="x", y="y")   # chiziqli (egri xato)   # ⚠️
sns.regplot(..., order=2)   # polinomial (egri)               # ✅

5. figure vs axes

sns.relplot(...); ax.set_title(...)   # relplot — figure (ax yo'q) # ⚠️
sns.scatterplot(..., ax=ax)   # axes (ax bilan)               # ✅

6. Eski API

sns.distplot(data)   # eski (o'chirilgan)                     # ⚠️
sns.histplot(data)   # yangi                                   # ✅

7. set_theme unutish

sns.histplot(...)   # standart Matplotlib ko'rinish           # ⚠️
sns.set_theme(); sns.histplot(...)   # chiroyli               # ✅

7. Integratsiya — bu bilim qayerda kerak bo'ladi

• 5.1-dars (o'tilgan): Matplotlib (poydevor)
• 3-qism (o'tilgan): Pandas (DataFrame)
• 5.8-dars: Heatmap (korrelyatsiya)
• 5.9-dars: Taqsimot grafiklari
• 9-qism (reja): EDA (tez vizualizatsiya)

8. Eng yaxshi amaliyotlar

1. sns.set_theme() — chiroyli uslub.

2. DataFrame — data=df, x=, y=.

3. hue= — guruh (kam guruh, rang).

4. Statistik — regplot/histplot/violinplot.

5. Bezash — Matplotlib (ax.set_title).

6. ax= — Matplotlib bilan birga.

7. regplot — chiziqli (nochiziqli — order=).

8. Seaborn — statistik vizualizatsiya kam kod.

9. Amaliy topshiriq

Vazifa 1: Bashorat qiling

1.  # Seaborn nima?
2.  # Matplotlib bilan aloqasi?
3.  # data= nima?
4.  # x=/y= nima?
5.  # hue nima?
6.  # regplot nima?
7.  # histplot kde?
8.  # violinplot nima?
9.  # bezash qayerda?
10. # hue ko'p guruh?
11. # Seaborn massivmi?
12. # nega Seaborn muhim?

Vazifa 2: Xatolarni tuzating

1.  sns.scatterplot(x=m1, y=m2)   # DataFrame

2.  sns.scatterplot(..., hue="id")   # 100 guruh

3.  sns.scatterplot(...); plt.title(...)

4.  sns.regplot(...)   # egri aloqa

5.  sns.distplot(data)   # eski

1.  data=df, x="a", y="b"

2.  hue="kategoriya" (kam guruh)

3.  ax=ax; ax.set_title(...)

4.  order=2 (polinomial)

5.  sns.histplot(data)

Vazifa 3: Seaborn

Modellang:

1. Matplotlib ustida
2. Chiroyli
3. Statistik
4. Kam kod

Vazifa 4: DataFrame

Modellang:

1. data=
2. x=/y=
3. hue=
4. Ustun nomlari

Vazifa 5: Statistik grafiklar

Modellang:

1. regplot
2. histplot (KDE)
3. violinplot
4. heatmap

Vazifa 6: Integratsiya

Modellang:

1. ax=
2. Bezash
3. Matplotlib
4. Birga

Vazifa 7: O'ylash

Seaborn Matplotlib ustida qurilgan — kam kod bilan chiroyli, statistik grafik beradi, lekin Matplotlib'ni almashtirmaydi (bezash uchun hali ham kerak). Nima uchun bu "qatlam ustida qatlam" naqshi (Pandas NumPy ustida — 3.1) Data Science ekotizimida takrorlanadi, va nega ikkalasini (yuqori + past daraja) bilish foydali?

Nimani mustahkamlaydi: 2.1, 2.5-bo'limlar.

Xulosa

Bu darsda Seaborn'ni o'rgandik.

Eng muhim uch fikr:

1. Seaborn va DataFrame. Seaborn (sns) — Matplotlib ustida qurilgan yuqori daraja (chiroyli, statistik, kam kod; sns.set_theme()). DataFrame bilan — sns.scatterplot(data=df, x="reklama", y="savdo", hue="shahar") (data= jadval, x=/y= ustun nomlari, hue= guruh — rang + avtomatik legend); Pandas bilan to'g'ridan.

2. Statistik grafiklar va hue. Statistik grafiklar (o'rnatilgan): regplot (scatter + regressiya chizig'i — trend), histplot (histogram + KDE silliq egri; kde=True), violinplot (box + shakl), heatmap (korrelyatsiya — 5.8), pairplot (barcha juftlik). hue (guruh — rang, avtomatik legend; barcha grafikda; kam guruh 2-6).

3. Kam kod, statistik. Matplotlib bilan — Seaborn chizadi, bezash/saqlash Matplotlib (ax=, ax.set_title, fig.savefig; birga). Seaborn — statistik vizualizatsiya kam kod bilan (Matplotlib ustida — chiroyli, statistik; DataFrame). Seaborn (yuqori — qulay) + Matplotlib (pastki — nazorat) — qatlam naqshi (Pandas NumPy ustida 3.1; ekotizim takrorlaydi; ikkalasi bil — yuqori tez, pastki maxsus). Data Science'da keng (EDA — tez chiroyli grafik). Tuzoqlar: DataFrame afzal (data=), hue ko'p guruh (chalkash), Matplotlib bezash, regplot chiziqli (nochiziqli order=), eski API (distplot → histplot).

Keyingi darsda heatmap (issiqlik xaritasi)ni o'rganamiz: korrelyatsiya matritsasini rangli ko'rsatish — ko'p o'zgaruvchan aloqasini bir qarashda (4.9 korrelyatsiya matritsasi).