Pandas — Series & DataFrame (Complete Notes)¶

CampusX Session 16, 17, 18 (Pandas Series + Pandas DataFrame + Important DataFrame Methods)¶

SESSION 16 — PANDAS SERIES¶

1. Pandas hai kya?¶

Pandas ek data analysis aur manipulation library hai. Ye NumPy ke upar bana hua hai.

NumPy vs Pandas:

NumPy	Pandas
Sirf ek datatype	Har column ka alag datatype ho sakta hai
Numeric index (0,1,2...)	Custom labels index bana sakte ho
Numerical computing	Real-world tabular data (Excel/CSV jaisa)
Missing values handle karna mushkil	Built-in missing value support

import pandas as pd
import numpy as np

Pandas ke do main objects: - Series → 1D labeled array (Excel ka ek column) - DataFrame → 2D labeled table (poori Excel sheet)

2. Series banana¶

2.1 List se¶

country = ['India', 'Pakistan', 'USA', 'Nepal', 'Sri Lanka']
pd.Series(country)
# 0        India
# 1     Pakistan
# 2          USA
# 3        Nepal
# 4    Sri Lanka
# dtype: object

2.2 Custom index ke saath¶

marks = [67, 57, 89, 100]
subjects = ['maths', 'english', 'science', 'hindi']

pd.Series(marks, index=subjects)
# maths       67
# english     57
# science     89
# hindi      100

2.3 `name` parameter¶

marks = pd.Series(marks, index=subjects, name='Nitish ke marks')

2.4 Dictionary se (sabse aasan)¶

marks = {'maths': 67, 'english': 57, 'science': 89, 'hindi': 100}
marks_series = pd.Series(marks, name='Nitish ke marks')

Dictionary ki keys → index ban jaati hain, values → data.

3. Series Attributes¶

marks_series.size        # kitne elements
marks_series.dtype       # datatype
marks_series.name        # series ka naam
marks_series.is_unique   # saari values unique hain?
marks_series.index       # index object
marks_series.values      # underlying NumPy array
marks_series.shape       # (4,)
marks_series.ndim        # 1

.index.is_unique se index unique hai ya nahi ye check karte hain.

4. Series `read_csv()` se banana¶

# Tareeqa 1: usecols + squeeze
subs = pd.read_csv('subs.csv').squeeze()

# Tareeqa 2: index_col ke saath
vk = pd.read_csv('kohli_ipl.csv', index_col='match_no').squeeze()

movies = pd.read_csv('bollywood.csv', index_col='movie').squeeze()

.squeeze() single-column DataFrame ko Series mein badal deta hai.

Purane code mein squeeze=True parameter hota tha, ab wo deprecated hai — .squeeze() method use karo.

5. Series Methods¶

5.1 `head()` aur `tail()`¶

subs.head()        # pehle 5
subs.head(3)       # pehle 3
subs.tail()        # aakhri 5
subs.tail(3)

5.2 `sample()`¶

movies.sample()     # random 1
movies.sample(5)    # random 5

5.3 `value_counts()` ⭐¶

Har unique value kitni baar aayi hai — descending order mein.

movies.value_counts()
# Akshay Kumar     48
# Amitabh Bachchan 45
# ...

movies.value_counts(normalize=True)   # percentage mein

5.4 `sort_values()` ⭐¶

vk.sort_values()                      # ascending
vk.sort_values(ascending=False)       # descending
vk.sort_values(ascending=False).head(1)          # highest score
vk.sort_values(ascending=False).head(1).values[0] # sirf value

vk.sort_values(ascending=False, inplace=True)    # original badal do

5.5 `sort_index()`¶

movies.sort_index()
movies.sort_index(ascending=False)

5.6 `inplace` parameter¶

inplace=True → original series modify ho jayegi, kuch return nahi hoga. inplace=False (default) → nayi series return hogi, original safe.

6. Series Math Methods¶

subs.count()        # non-null values ki count (NaN nahi ginta)
subs.sum()
subs.product()
subs.mean()
subs.median()
subs.mode()
subs.std()
subs.var()
subs.min()
subs.max()
subs.describe()     # sab ek saath: count, mean, std, min, 25%, 50%, 75%, max

⚠️ Note: len() saare elements ginta hai, count() sirf non-null.

7. Series with Python Functionality¶

7.1 Built-in functions¶

len(marks_series)          # length
type(marks_series)         # pandas.core.series.Series
dir(marks_series)          # saare available methods
sorted(marks_series)       # sorted list return karega
min(marks_series)
max(marks_series)

7.2 Type conversion¶

list(marks_series)         # list ban jayegi (sirf values)
dict(marks_series)         # dictionary (index → value)

7.3 Membership operator (`in`) ⚠️¶

'maths' in marks_series      # True  → INDEX check karta hai
67 in marks_series           # False → values check NAHI karta!

67 in marks_series.values    # True  → values check karne ka sahi tareeqa

Ye trap hai — in hamesha index pe kaam karta hai.

7.4 Looping¶

for i in marks_series:
    print(i)                 # VALUES print hongi

for i in marks_series.index:
    print(i)                 # INDEX print hoga

for index, value in marks_series.items():
    print(index, value)      # dono

7.5 Arithmetic operators (Broadcasting)¶

100 - marks_series           # har value ko 100 se minus
marks_series + 10
marks_series * 2

7.6 Relational operators¶

vk >= 50                     # boolean series

8. Boolean Indexing on Series ⭐⭐¶

Ye Pandas ka sabse powerful feature hai.

# Kohli ne kitni baar 50+ score kiya?
(vk >= 50).sum()

# Kitni baar duck out hua (0 runs)?
(vk == 0).sum()

# Kitne din mein 200 se zyada subscribers mile?
(subs > 200).sum()

# Actual values dekho
vk[vk >= 50]

Pattern: series[condition] → sirf wo values jahan condition True hai.

Multiple conditions:

vk[(vk >= 50) & (vk < 100)]     # fifties (not centuries)
vk[(vk == 0) | (vk == 1)]

and/or/not ki jagah &, |, ~ — aur har condition brackets mein.

9. Plotting on Series¶

subs.plot()                          # line plot (time series ke liye)
movies.value_counts().head(20).plot(kind='bar')
movies.value_counts().head(10).plot(kind='pie')
vk.plot(kind='hist')

10. Important Series Methods (Supplementary Session)¶

10.1 `astype()` — memory optimization ⭐¶

import sys
sys.getsizeof(vk)                       # pehle
vk = vk.astype('int16')
sys.getsizeof(vk)                       # ab bahut kam

10.2 `between()`¶

vk[vk.between(51, 99)]                  # 51 se 99 ke beech (dono include)
vk[vk.between(51, 99)].size             # kitni fifties

10.3 `clip()`¶

subs.clip(100, 200)     # 100 se kam → 100, 200 se zyada → 200

10.4 `drop_duplicates()`¶

temp = pd.Series([1, 1, 2, 2, 3, 3, 4, 4])
temp.drop_duplicates()                  # pehla wala rakhta hai
temp.drop_duplicates(keep='last')       # aakhri wala rakhta hai
temp.duplicated()                       # boolean: duplicate hai ya nahi
temp.duplicated().sum()                 # kitne duplicates

10.5 `isnull()` / `dropna()` / `fillna()`¶

temp = pd.Series([1, 2, 3, np.nan, 5, 6, np.nan, 8, np.nan, 10])

temp.isnull()          # boolean
temp.isnull().sum()    # kitne missing → 3
temp.count()           # non-null count → 7
temp.size              # total → 10

temp.dropna()          # missing rows hata do
temp.fillna(0)         # 0 se bharo
temp.fillna(temp.mean())   # mean se bharo ⭐ (common technique)
temp.fillna(method='ffill')  # pichli value se bharo
temp.fillna(method='bfill')  # agli value se bharo

10.6 `isin()`¶

vk[vk.isin([49, 99])]     # sirf 49 aur 99 wale scores

10.7 `apply()` ⭐¶

Custom function har element pe apply karta hai.

# Har movie ke actor ka sirf pehla naam
movies.apply(lambda x: x.split()[0].upper())

# Subscribers ko 'good day' / 'bad day' mein classify karo
subs.apply(lambda x: 'good day' if x > subs.mean() else 'bad day')

10.8 `copy()` — View vs Copy ⚠️¶

new = vk.head()          # ye ek VIEW hai
new[1] = 100             # original vk BHI badal jayega!

new = vk.head().copy()   # ye ek COPY hai
new[1] = 100             # original safe hai

---¶

SESSION 17 — PANDAS DATAFRAME¶

1. DataFrame kya hai?¶

2D labeled data structure — rows aur columns dono ke labels hote hain. Excel sheet ki tarah socho.

DataFrame = Series ka collection (har column ek Series hai).

2. DataFrame banana¶

2.1 Lists se¶

student_data = [
    [100, 80, 10],
    [90, 70, 7],
    [120, 100, 14],
    [80, 50, 2]
]
pd.DataFrame(student_data, columns=['iq', 'marks', 'package'])

2.2 Dictionary se¶

student_dict = {
    'name': ['nitish', 'ankit', 'rupesh', 'rishabh', 'amit', 'ankita'],
    'iq': [100, 90, 120, 80, 0, 0],
    'marks': [80, 70, 100, 50, 0, 0],
    'package': [10, 7, 14, 2, 0, 0]
}
students = pd.DataFrame(student_dict)
students.set_index('name', inplace=True)

2.3 CSV se ⭐ (real world mein yahi use hota hai)¶

movies = pd.read_csv('movies.csv')
ipl = pd.read_csv('ipl-matches.csv')

# Useful parameters:
pd.read_csv('file.csv', index_col='name')
pd.read_csv('file.csv', usecols=['col1', 'col2'])
pd.read_csv('file.csv', nrows=100)
pd.read_csv('file.csv', sep='\t')          # tab separated
pd.read_csv('file.csv', encoding='latin-1')
pd.read_csv('file.csv', header=None)
pd.read_csv('file.csv', na_values=['?', '-'])
pd.read_csv('file.csv', parse_dates=['date_col'])

3. DataFrame Attributes¶

movies.shape        # (rows, columns)
movies.dtypes       # har column ka datatype
movies.index        # row labels
movies.columns      # column names
movies.values       # NumPy 2D array
movies.size         # total cells
movies.ndim         # 2

4. DataFrame Methods¶

4.1 `head()` / `tail()` / `sample()`¶

movies.head(2)
movies.tail(3)
ipl.sample(5)

4.2 `info()` ⭐¶

movies.info()

Ye batata hai: column names, non-null count, dtype, memory usage. Sabse pehle yahi chalao kisi naye dataset pe.

4.3 `describe()` ⭐¶

movies.describe()      # sirf numerical columns ka stats

count, mean, std, min, 25%, 50%, 75%, max.

4.4 `isnull()`¶

movies.isnull()             # boolean DataFrame
movies.isnull().sum()       # har column mein kitne missing ⭐

4.5 `duplicated()`¶

movies.duplicated().sum()   # kitni duplicate rows
students.drop_duplicates()

4.6 `rename()`¶

students.rename(columns={'marks': 'percent', 'package': 'lpa'}, inplace=True)
students.rename(index={'nitish': 'Nitish', 'ankit': 'Ankit'}, inplace=True)

5. DataFrame Math Methods¶

students.sum()             # column-wise (axis=0 default)
students.sum(axis=1)       # row-wise
students.mean()
students.median()
students.mode()
students.std()
students.var()
students.min()
students.max()

6. Selecting Columns aur Rows ⭐⭐¶

6.1 Single column (Series milegi)¶

movies['title_x']
movies.title_x         # dot notation (space wale names mein kaam nahi karega)

6.2 Multiple columns (DataFrame milega)¶

movies[['year_of_release', 'actors', 'title_x']]

Note: Double brackets [[ ]].

6.3 Rows select karna — `iloc` aur `loc` ⭐⭐⭐¶

	`iloc`	`loc`
Kaam	Position (integer) se	Label se
Slicing	End exclude hota hai	End include hota hai
Example	`df.iloc[0:5]` → 0,1,2,3,4	`df.loc['a':'e']` → a,b,c,d,e

# iloc — integer position
movies.iloc[0]              # pehli row
movies.iloc[0:5]            # pehli 5 rows
movies.iloc[[0, 4, 5]]      # fancy indexing
movies.iloc[::2]            # har doosri row

# loc — label based
students.loc['nitish']
students.loc['nitish':'rishabh']       # rishabh BHI include hoga
students.loc[['nitish', 'ankita', 'rupesh']]
students.loc['nitish':'rishabh':2]

6.4 Rows aur Columns dono ek saath¶

movies.iloc[0:3, 0:3]                          # position se
movies.loc[0:2, 'title_x':'poster_path']       # label se

students.iloc[0:3, 1:]
students.loc['nitish':'rishabh', 'iq':'marks']

# Mix karna ho toh:
movies.iloc[0:3][['title_x', 'actors']]

7. Filtering a DataFrame ⭐⭐⭐¶

Ye asli data analysis hai. Pattern: df[condition]

ipl = pd.read_csv('ipl-matches.csv')

Example 1: Final matches¶

mask = ipl['MatchNumber'] == 'Final'
ipl[mask]
ipl[mask][['Season', 'WinningTeam']]      # sirf 2 columns

Example 2: Ek team ne kitne match jeete¶

ipl[ipl['WinningTeam'] == 'Chennai Super Kings'].shape[0]

Example 3: Toss jeeta AUR match bhi jeeta (AND)¶

ipl[(ipl['TossWinner'] == ipl['WinningTeam'])].shape[0]

Example 4: Multiple conditions¶

# 200+ runs se jeetne wale matches
ipl[(ipl['WonBy'] == 'Runs') & (ipl['Margin'] > 100)]

# Super 8 ya Final
ipl[ipl['MatchNumber'].isin(['Final', 'Qualifier 1'])]

Example 5: Movies dataset¶

# 8+ rating AND 10000+ votes
movies[(movies['imdb_rating'] > 8) & (movies['no_of_votes'] > 10000)]

# Action movies with 7.5+ rating
mask1 = movies['genres'].str.contains('Action')
mask2 = movies['imdb_rating'] > 7.5
movies[mask1 & mask2]

Zaroori operators:

&    # AND
|    # OR
~    # NOT
.isin([...])            # multiple values mein se koi ek
.str.contains('text')   # text search
.between(a, b)          # range

8. Naye Columns add karna¶

8.1 Bilkul naya column¶

movies['Country'] = 'India'

8.2 Existing columns se derive karke ⭐¶

students['percentage'] = (students['marks'] / 100) * 100

movies['lead_actor'] = movies['actors'].str.split('|').apply(lambda x: x[0])

ipl['margin_type'] = np.where(ipl['WonBy'] == 'Runs', 'Batting First', 'Chasing')

8.3 `insert()` — specific position pe column¶

students.insert(loc=1, column='gender', value=['M','M','M','M','F','F'])

9. `astype()` — Memory Optimization ⭐¶

ipl.info()                         # pehle memory dekho

ipl['Season'] = ipl['Season'].astype('category')
ipl['Team1'] = ipl['Team1'].astype('category')
ipl['Team2'] = ipl['Team2'].astype('category')
ipl['ID'] = ipl['ID'].astype('int32')

ipl.info()                         # ab memory bahut kam

Rule of thumb: Jin columns mein kam unique values hain (jaise city names, categories), unko category bana do. Numbers ko sabse chhote possible int/float type mein daalo.

---¶

SESSION 18 — IMPORTANT DATAFRAME METHODS¶

Setup:

movies = pd.read_csv('imdb-top-1000.csv')
ipl = pd.read_csv('ipl-matches.csv')
batsman = pd.read_csv('batsman_runs_ipl.csv')

1. `value_counts()` ⭐¶

Series pe¶

marks.value_counts()

DataFrame pe (multiple columns ka combination)¶

ipl[['Team1', 'Team2', 'WinningTeam']].value_counts()

Real use cases:

# Sabse zyada player of the match jeetne wale
ipl['Player_of_Match'].value_counts().head(10)

# Toss decision ka breakdown
ipl['TossDecision'].value_counts()

# Har season mein kitne match
ipl['Season'].value_counts()

2. `sort_values()` ⭐¶

movies.sort_values('title_x')                       # ascending
movies.sort_values('title_x', ascending=False)      # descending

# Multiple columns
movies.sort_values(['year_of_release', 'title_x'], ascending=[True, False])

# NaN ko upar/neeche rakhna
movies.sort_values('imdb_rating', na_position='first')

movies.sort_values('imdb_rating', inplace=True)

3. `rank()`¶

batsman['batting_rank'] = batsman['batsman_run'].rank(ascending=False)
batsman.sort_values('batting_rank')

Ranking assign karta hai. Ties ke liye average rank deta hai by default.

4. `sort_index()`¶

marks_series.sort_index(ascending=False)
movies.sort_index()

5. `set_index()` aur `reset_index()` ⭐¶

batsman.set_index('batter', inplace=True)      # column ko index bana do
batsman.reset_index(inplace=True)              # index ko wapas column bana do

# Index ko bilkul hata dena
batsman.reset_index(drop=True, inplace=True)

Use case: value_counts() ka result Series hota hai; reset_index() se DataFrame bana lete hain.

ipl['Player_of_Match'].value_counts().reset_index()

6. `rename()`¶

movies.rename(columns={'imdb_id': 'imdb', 'poster_path': 'link'}, inplace=True)
movies.rename(index={0: 'zero', 1: 'one'}, inplace=True)

7. `unique()` aur `nunique()` ⭐¶

temp = pd.Series([1, 1, 2, 2, 3, 3, 4, 4, np.nan, np.nan])

temp.unique()         # array with NaN included
temp.nunique()        # 4 (NaN ko count NAHI karta)
temp.nunique(dropna=False)   # 5

ipl['Season'].unique()
ipl['Season'].nunique()      # kitne seasons hue

Farq: unique() values deta hai, nunique() count deta hai. unique() NaN ginta hai, nunique() nahi.

8. `isnull()` / `notnull()` / `hasnans`¶

students['name'].isnull()      # missing hai
students['name'].notnull()     # missing nahi hai
students['name'].hasnans       # koi missing hai kya? (True/False)

movies.isnull().sum()          # har column ka missing count ⭐
movies.isnull().sum().sum()    # total missing

9. `dropna()` ⭐¶

students.dropna()                          # jis row mein bhi NaN ho, hata do
students.dropna(how='all')                 # sirf jab POORI row NaN ho
students.dropna(how='any')                 # default
students.dropna(subset=['name'])           # sirf 'name' column dekho
students.dropna(subset=['name', 'iq'])     # in dono mein se koi bhi NaN ho
students.dropna(axis=1)                    # NaN wale COLUMNS hata do
students.dropna(thresh=2)                  # kam se kam 2 non-null hone chahiye

10. `fillna()` ⭐¶

students['name'].fillna('unknown')
students['package'].fillna(students['package'].mean())      # mean imputation
students['package'].fillna(students['package'].median())    # median (outliers ho toh better)
students['name'].fillna(students['name'].mode()[0])         # mode (categorical)

students.fillna(method='ffill')     # forward fill (pichli value)
students.fillna(method='bfill')     # backward fill (agli value)
students.fillna(0)

11. `drop_duplicates()` ⭐¶

marks.drop_duplicates()
marks.drop_duplicates(keep='last')
marks.drop_duplicates(subset=['name'])       # sirf name column dekho

# Classic question: har IPL final ka winner
ipl.drop_duplicates(subset=['Season'], keep='last')[['Season', 'WinningTeam']]

duplicated() boolean deta hai:

students.duplicated().sum()

12. `drop()` ⭐¶

# Columns hatao
temp.drop(columns=['branch', 'cgpa'])
temp.drop(columns=['branch'], inplace=True)

# Rows hatao
temp.drop(index=[0, 1, 2])
students.drop(index=['nitish', 'aditya'])

# axis se
temp.drop('branch', axis=1)     # column
temp.drop(0, axis=0)            # row

13. `apply()` ⭐⭐¶

Series pe¶

points_df['distance'] = points_df['2d_point'].apply(euclidean)

# Simple lambda
movies['lead_actor'] = movies['actors'].apply(lambda x: x.split('|')[0])

DataFrame pe (axis matter karta hai)¶

def euclidean(row):
    pt_A = row['1st point']
    pt_B = row['2nd point']
    return ((pt_A[0] - pt_B[0])**2 + (pt_A[1] - pt_B[1])**2) ** 0.5

points_df['distance'] = points_df.apply(euclidean, axis=1)

axis=1 → function ko har row milegi. axis=0 → function ko har column milega.

14. `isin()`¶

ipl[ipl['Season'].isin(['2007/08', '2009', '2020/21'])]
students[students['branch'].isin(['EEE', 'ECE'])]

15. `corr()` — correlation¶

movies.corr(numeric_only=True)
movies.corr(numeric_only=True)['imdb_rating'].sort_values(ascending=False)

-1 se 1 tak. 1 = perfect positive, -1 = perfect negative, 0 = koi relation nahi.

16. `nlargest()` aur `nsmallest()`¶

movies.nlargest(5, 'imdb_rating')
movies.nsmallest(5, 'imdb_rating')
students.nlargest(3, 'package')

sort_values().head(n) se better hai — zyada efficient.

17. `insert()`¶

movies.insert(loc=0, column='rank', value=range(1, len(movies)+1))

18. `copy()` — View vs Copy ⚠️¶

new = movies.head()          # VIEW — original se juda
new = movies.head().copy()   # COPY — independent

Agar copy nahi banaya aur modify kiya toh SettingWithCopyWarning aata hai.

19. Extra useful methods¶

df.replace({'old': 'new'})
df.astype({'col': 'int32'})
df.round(2)
df.abs()
df.cumsum()
df.diff()                    # consecutive difference
df.pct_change()              # percentage change
df.shift(1)                  # values ko shift karo
df.clip(lower, upper)
df.memory_usage(deep=True)
df.select_dtypes(include='number')
df.columns.tolist()
df.T                         # transpose
df.to_csv('output.csv', index=False)
df.to_excel('output.xlsx')

Session 16-18 Quick Revision¶

Kaam	Method
Data load karo	`read_csv`, `read_excel`
Pehli nazar	`head`, `tail`, `sample`, `info`, `describe`, `shape`
Column select	`df['col']`, `df[['a','b']]`
Row select	`iloc` (position), `loc` (label)
Filter	`df[mask]`, `&`, `\\|`, `~`, `isin`, `str.contains`, `between`
Sort	`sort_values`, `sort_index`, `nlargest`, `nsmallest`, `rank`
Missing values	`isnull`, `dropna`, `fillna`
Duplicates	`duplicated`, `drop_duplicates`
Naya column	`df['new'] = ...`, `insert`, `apply`
Frequency	`value_counts`, `unique`, `nunique`
Index	`set_index`, `reset_index`, `rename`
Memory	`astype`, `memory_usage`

Pandas — Series & DataFrame (Complete Notes)¶

CampusX Session 16, 17, 18 (Pandas Series + Pandas DataFrame + Important DataFrame Methods)¶

SESSION 16 — PANDAS SERIES¶

1. Pandas hai kya?¶

2. Series banana¶

2.1 List se¶

2.2 Custom index ke saath¶

2.3 name parameter¶

2.4 Dictionary se (sabse aasan)¶

3. Series Attributes¶

4. Series read_csv() se banana¶

5. Series Methods¶

5.1 head() aur tail()¶

5.2 sample()¶

5.3 value_counts() ⭐¶

5.4 sort_values() ⭐¶

5.5 sort_index()¶

5.6 inplace parameter¶

6. Series Math Methods¶

7. Series with Python Functionality¶

7.1 Built-in functions¶

7.2 Type conversion¶

7.3 Membership operator (in) ⚠️¶

7.4 Looping¶

7.5 Arithmetic operators (Broadcasting)¶

7.6 Relational operators¶

8. Boolean Indexing on Series ⭐⭐¶

9. Plotting on Series¶

10. Important Series Methods (Supplementary Session)¶

10.1 astype() — memory optimization ⭐¶

10.2 between()¶

10.3 clip()¶

10.4 drop_duplicates()¶

10.5 isnull() / dropna() / fillna()¶

10.6 isin()¶

10.7 apply() ⭐¶

10.8 copy() — View vs Copy ⚠️¶

---¶

SESSION 17 — PANDAS DATAFRAME¶

1. DataFrame kya hai?¶

2. DataFrame banana¶

2.1 Lists se¶

2.2 Dictionary se¶

2.3 CSV se ⭐ (real world mein yahi use hota hai)¶

3. DataFrame Attributes¶

4. DataFrame Methods¶

4.1 head() / tail() / sample()¶

4.2 info() ⭐¶

4.3 describe() ⭐¶

4.4 isnull()¶

4.5 duplicated()¶

4.6 rename()¶

5. DataFrame Math Methods¶

6. Selecting Columns aur Rows ⭐⭐¶

6.1 Single column (Series milegi)¶

6.2 Multiple columns (DataFrame milega)¶

6.3 Rows select karna — iloc aur loc ⭐⭐⭐¶

6.4 Rows aur Columns dono ek saath¶

7. Filtering a DataFrame ⭐⭐⭐¶

Example 1: Final matches¶

Example 2: Ek team ne kitne match jeete¶

Example 3: Toss jeeta AUR match bhi jeeta (AND)¶

Example 4: Multiple conditions¶

Example 5: Movies dataset¶

8. Naye Columns add karna¶

8.1 Bilkul naya column¶

8.2 Existing columns se derive karke ⭐¶

8.3 insert() — specific position pe column¶

9. astype() — Memory Optimization ⭐¶

---¶

SESSION 18 — IMPORTANT DATAFRAME METHODS¶

1. value_counts() ⭐¶

Series pe¶

DataFrame pe (multiple columns ka combination)¶

2. sort_values() ⭐¶

3. rank()¶

4. sort_index()¶

5. set_index() aur reset_index() ⭐¶

6. rename()¶

7. unique() aur nunique() ⭐¶

2.3 `name` parameter¶

4. Series `read_csv()` se banana¶

5.1 `head()` aur `tail()`¶

5.2 `sample()`¶

5.3 `value_counts()` ⭐¶

5.4 `sort_values()` ⭐¶

5.5 `sort_index()`¶

5.6 `inplace` parameter¶

7.3 Membership operator (`in`) ⚠️¶

10.1 `astype()` — memory optimization ⭐¶

10.2 `between()`¶

10.3 `clip()`¶

10.4 `drop_duplicates()`¶

10.5 `isnull()` / `dropna()` / `fillna()`¶

10.6 `isin()`¶

10.7 `apply()` ⭐¶

10.8 `copy()` — View vs Copy ⚠️¶

4.1 `head()` / `tail()` / `sample()`¶

4.2 `info()` ⭐¶

4.3 `describe()` ⭐¶

4.4 `isnull()`¶

4.5 `duplicated()`¶

4.6 `rename()`¶

6.3 Rows select karna — `iloc` aur `loc` ⭐⭐⭐¶

8.3 `insert()` — specific position pe column¶

9. `astype()` — Memory Optimization ⭐¶

1. `value_counts()` ⭐¶

2. `sort_values()` ⭐¶

3. `rank()`¶

4. `sort_index()`¶

5. `set_index()` aur `reset_index()` ⭐¶

6. `rename()`¶

7. `unique()` aur `nunique()` ⭐¶

8. `isnull()` / `notnull()` / `hasnans`¶

9. `dropna()` ⭐¶

10. `fillna()` ⭐¶

11. `drop_duplicates()` ⭐¶

12. `drop()` ⭐¶

13. `apply()` ⭐⭐¶

14. `isin()`¶

15. `corr()` — correlation¶

16. `nlargest()` aur `nsmallest()`¶

17. `insert()`¶

18. `copy()` — View vs Copy ⚠️¶