24.6-dars: pandas — o'qish va yozish

import pandas as pd
df = pd.read_csv("mahsulot.csv")
# CSV:
# nom,narx,til
# Python,100,uz
# Go,120,en

df.to_csv("natija.csv", index=False)
# index=False — indeks ustunini yozmaydi
df.to_csv("natija.csv", index=False, encoding="utf-8")

pd.read_csv("f.csv", sep=";")           # nuqta-vergul ajratgich
pd.read_csv("f.tsv", sep="\t")          # tab (TSV)
pd.read_csv("f.csv", header=None)        # sarlavha yo'q
pd.read_csv("f.csv", names=["a", "b"])   # ustun nomlari
pd.read_csv("f.csv", usecols=["nom"])    # faqat ba'zi ustun
pd.read_csv("f.csv", nrows=100)          # faqat 100 qator

pd.read_csv("f.csv", parse_dates=["sana"])
# "2024-01-15" (matn) → sana obyekti (datetime)
df["sana"].dt.year        # sanadan yil
df["sana"].dt.month       # oy

df.isna()               # qayerda bo'sh (bool)
df.isna().sum()         # har ustunda bo'sh soni
df.dropna()             # bo'sh qatorlarni o'chir
df.fillna(0)            # bo'shni 0 bilan to'ldir
pd.read_csv("f.csv", na_values=["N/A", "-"])  # bo'sh belgilari

df = pd.read_json("data.json")
df.to_json("out.json", orient="records")
# orient="records": [{"nom": "Python", ...}, {...}]
df.to_dict(orient="records")   # Python ro'yxatiga

df = pd.read_excel("hisobot.xlsx")            # o'qi (openpyxl kerak)
df = pd.read_excel("f.xlsx", sheet_name="Sotuv")  # varaq
df.to_excel("natija.xlsx", index=False)       # yoz
# bir necha varaq:
with pd.ExcelWriter("out.xlsx") as w:
    df1.to_excel(w, sheet_name="Q1")
    df2.to_excel(w, sheet_name="Q2")

import pandas as pd

# CSV:
df = pd.read_csv("f.csv")
df = pd.read_csv("f.csv", sep=";", parse_dates=["sana"], na_values=["N/A"])
df.to_csv("out.csv", index=False)

# JSON:
df = pd.read_json("f.json")
df.to_json("out.json", orient="records")
df.to_dict(orient="records")

# Excel (openpyxl):
df = pd.read_excel("f.xlsx", sheet_name="Sotuv")
df.to_excel("out.xlsx", index=False)

# bo'sh qiymatlar:
df.isna().sum()             # har ustunda bo'sh soni
df.dropna(); df.fillna(0)

# StringIO (sinash uchun):
from io import StringIO
pd.read_csv(StringIO("a,b\n1,2\n"))

read_csv/to_csv (index=False) · read_json/to_json (orient=records)
read_excel/to_excel (openpyxl) · parse_dates: sana · na_values: bo'sh
ETL: o'qi → tozala → yoz

"""read_csv (StringIO'dan); to_csv (index=False — indekssiz); ustun turlar avtomatik."""

import warnings
warnings.filterwarnings("ignore")

from io import StringIO

import pandas as pd


def main() -> None:
    csv_matn = "nom,narx,til\nPython,100,uz\nGo,120,en\nRust,150,en\n"

    print("=== 1. read_csv (o'qish) ===")
    df = pd.read_csv(StringIO(csv_matn))
    print(df.to_string(index=False))

    print("\n=== 2. Turlar avtomatik aniqlandi ===")
    print(f"  dtypes: {df.dtypes.astype(str).to_dict()}")

    print("\n=== 3. to_csv (yozish, index=False) ===")
    natija = df.to_csv(index=False)
    print(natija.strip())

    print("\n=== 4. Statistika (o'qilgan ma'lumot) ===")
    print(f"  o'rtacha narx: {round(df['narx'].mean(), 1)}")
    print(f"  qatorlar: {len(df)}")
    print("  ⭐ read_csv/to_csv — CSV o'qish va yozish")


if __name__ == "__main__":
    main()

=== 1. read_csv (o'qish) ===
   nom  narx til
Python   100  uz
    Go   120  en
  Rust   150  en

=== 2. Turlar avtomatik aniqlandi ===
  dtypes: {'nom': 'str', 'narx': 'int64', 'til': 'str'}

=== 3. to_csv (yozish, index=False) ===
nom,narx,til

Python,100,uz

Go,120,en

Rust,150,en

=== 4. Statistika (o'qilgan ma'lumot) ===
  o'rtacha narx: 123.3
  qatorlar: 3
  ⭐ read_csv/to_csv — CSV o'qish va yozish

"""o'qish sozlamalari: sep (ajratgich), parse_dates (sana turi), na_values (bo'sh belgi)."""

import warnings
warnings.filterwarnings("ignore")

from io import StringIO

import pandas as pd


def main() -> None:
    print("=== 1. sep=';' (nuqta-vergul) ===")
    matn = "nom;narx\nPython;100\nGo;120\n"
    df = pd.read_csv(StringIO(matn), sep=";")
    print(f"  ustunlar: {list(df.columns)}, narxlar: {df['narx'].tolist()}")

    print("\n=== 2. parse_dates (sana turi) ===")
    matn2 = "sana,summa\n2024-01-01,100\n2024-02-01,200\n2024-03-01,150\n"
    df2 = pd.read_csv(StringIO(matn2), parse_dates=["sana"])
    print(f"  sana turi: {df2['sana'].dtype}")
    print(f"  yillar: {df2['sana'].dt.year.tolist()}, oylar: {df2['sana'].dt.month.tolist()}")

    print("\n=== 3. na_values (bo'sh belgi) ===")
    matn3 = "nom,narx\nPython,100\nGo,N/A\nRust,150\n"
    df3 = pd.read_csv(StringIO(matn3), na_values=["N/A"])
    print(f"  bo'sh soni: {df3['narx'].isna().sum()}")

    print("\n=== 4. usecols (faqat ba'zi ustun) ===")
    matn4 = "a,b,c\n1,2,3\n4,5,6\n"
    df4 = pd.read_csv(StringIO(matn4), usecols=["a", "c"])
    print(f"  tanlangan ustunlar: {list(df4.columns)}")
    print("  ⭐ sozlamalar — real CSV formatiga moslash")


if __name__ == "__main__":
    main()

=== 1. sep=';' (nuqta-vergul) ===
  ustunlar: ['nom', 'narx'], narxlar: [100, 120]

=== 2. parse_dates (sana turi) ===
  sana turi: datetime64[us]
  yillar: [2024, 2024, 2024], oylar: [1, 2, 3]

=== 3. na_values (bo'sh belgi) ===
  bo'sh soni: 1

=== 4. usecols (faqat ba'zi ustun) ===
  tanlangan ustunlar: ['a', 'c']
  ⭐ sozlamalar — real CSV formatiga moslash

"""JSON o'qish/yozish (orient=records); isna/fillna/dropna (bo'sh qiymatlar)."""

import warnings
warnings.filterwarnings("ignore")

from io import StringIO

import pandas as pd


def main() -> None:
    df = pd.DataFrame({
        "nom": ["Python", "Go", "Rust"],
        "narx": [100, 120, 150],
    })

    print("=== 1. to_json (orient=records) ===")
    j = df.to_json(orient="records")
    print(f"  {j}")

    print("\n=== 2. read_json (qayta o'qish) ===")
    df2 = pd.read_json(StringIO(j))
    print(f"  shape: {df2.shape}, nomlar: {df2['nom'].tolist()}")

    print("\n=== 3. Bo'sh qiymatlar (isna) ===")
    matn = "x,y\n1,\n,4\n5,6\n"
    d = pd.read_csv(StringIO(matn))
    print(f"  har ustunda bo'sh: {d.isna().sum().to_dict()}")

    print("\n=== 4. fillna va dropna ===")
    print(f"  fillna(0): {d.fillna(0)['x'].tolist()}")
    print(f"  dropna qatorlar: {len(d.dropna())}")
    print("  ⭐ JSON (API), bo'sh qiymatlar (real ma'lumot)")


if __name__ == "__main__":
    main()

=== 1. to_json (orient=records) ===
  [{"nom":"Python","narx":100},{"nom":"Go","narx":120},{"nom":"Rust","narx":150}]

=== 2. read_json (qayta o'qish) ===
  shape: (3, 2), nomlar: ['Python', 'Go', 'Rust']

=== 3. Bo'sh qiymatlar (isna) ===
  har ustunda bo'sh: {'x': 1, 'y': 1}

=== 4. fillna va dropna ===
  fillna(0): [1.0, 0.0, 5.0]
  dropna qatorlar: 1
  ⭐ JSON (API), bo'sh qiymatlar (real ma'lumot)

"""to'liq ETL: ikki CSV o'qi (concat), bo'sh tozala, hisob-kitob, natijani CSV'ga yoz."""

import warnings
warnings.filterwarnings("ignore")

from io import StringIO

import pandas as pd


def main() -> None:
    # ikki bo'lim CSV'si
    yanvar = "mahsulot,sotildi\nPython,50\nGo,30\nRust,20\n"
    fevral = "mahsulot,sotildi\nPython,60\nGo,\nJava,40\n"    # Go bo'sh

    print("=== 1. Ikki CSV o'qi va birlashtir ===")
    df1 = pd.read_csv(StringIO(yanvar))
    df1["oy"] = "yanvar"
    df2 = pd.read_csv(StringIO(fevral))
    df2["oy"] = "fevral"
    df = pd.concat([df1, df2], ignore_index=True)
    print(f"  jami qatorlar: {len(df)}")

    print("\n=== 2. Bo'sh qiymatlarni tekshir va tozala ===")
    print(f"  bo'sh sotildi: {df['sotildi'].isna().sum()}")
    df["sotildi"] = df["sotildi"].fillna(0).astype(int)
    print(f"  tozalangandan keyin: {df['sotildi'].tolist()}")

    print("\n=== 3. Hisob-kitob (mahsulot bo'yicha jami) ===")
    jami = df.groupby("mahsulot")["sotildi"].sum().sort_index()
    print(f"  {jami.to_dict()}")

    print("\n=== 4. Natijani yoz (CSV) ===")
    natija = jami.reset_index().to_csv(index=False)
    print(natija.strip())
    print("  ⭐ ETL — o'qi (concat) → tozala (fillna) → yoz (to_csv)")


if __name__ == "__main__":
    main()

=== 1. Ikki CSV o'qi va birlashtir ===
  jami qatorlar: 6

=== 2. Bo'sh qiymatlarni tekshir va tozala ===
  bo'sh sotildi: 1
  tozalangandan keyin: [50, 30, 20, 60, 0, 40]

=== 3. Hisob-kitob (mahsulot bo'yicha jami) ===
  {'Go': 30, 'Java': 40, 'Python': 110, 'Rust': 20}

=== 4. Natijani yoz (CSV) ===
mahsulot,sotildi

Go,30

Java,40

Python,110

Rust,20
  ⭐ ETL — o'qi (concat) → tozala (fillna) → yoz (to_csv)

df.to_csv("f.csv")               # ⚠️ indeks ustuni qo'shiladi
df.to_csv("f.csv", index=False)  # ✅

pd.read_csv("f.csv")   # ; bilan — bir ustun bo'ladi  # ⚠️
pd.read_csv("f.csv", sep=";")                          # ✅

pd.read_csv("f.csv")   # sana — matn                   # ⚠️
pd.read_csv("f.csv", parse_dates=["sana"])             # ✅

df["narx"].mean()  # NaN bo'lsa noto'g'ri              # ⚠️
df["narx"].isna().sum()  # avval tekshir                # ✅

pd.read_csv("f.csv")   # ba'zan UnicodeError            # ⚠️
pd.read_csv("f.csv", encoding="utf-8")                  # ✅

pd.read_excel("f.xlsx")   # ImportError                 # ⚠️
# pip install openpyxl                                   # ✅

pd.read_csv("katta.csv")   # xotira to'ladi             # ⚠️
pd.read_csv("katta.csv", nrows=1000)  # yoki chunksize   # ✅

1.  # read_csv nima?
2.  # to_csv nima?
3.  # index=False nima?
4.  # sep nima?
5.  # parse_dates nima?
6.  # na_values nima?
7.  # isna nima?
8.  # fillna nima?
9.  # dropna nima?
10. # to_json orient nima?
11. # read_excel nima kerak?
12. # ETL nima?

1.  df.to_csv("f.csv")             # index=False

2.  pd.read_csv("f.csv")  # ; bilan  # sep

3.  pd.read_csv("f.csv")  # sana     # parse_dates

4.  df["narx"].mean()  # NaN bor     # isna tekshir

5.  pd.read_csv("katta.csv")         # nrows