pre_machine learning

ml libraries
-pandas
reading dataset

df=pd.read_csv("address")
df=pd.read_excel("address",sheet_name="sheet",usecols=['coloumn1'])

viewing dataset

print(df.head)
print(df.tail)
print(df.sample(5))

daatbase info

db.info() #guves full information
db.describe()  #gives satistics
db.shape 

selecting coloumns

coloumn=df["coloumn_name"]
subset=df[["coloumn1","coloumn2"]]

addtwo compatable coloumns

df["new"]=df["col1"]+df["col2"]

filterrows

coloumn=df[df['coloumn']>10]

mergetwo dataframe

newdf=pd.merge(df1,df2,on="commen_coloumn")

missing values
-check

values=df.isnull().sum()

-fill with values

df['col'].filna(0,implace=True)

-drop

df=df.dropna()

-duplicate

df.df.duplicate().sum()

add new coloumn

y=np.array([1,2,3,4])
df['col_new']=y

drop col

df.drop(coloummns=['col_name'],implace=True)

ilock function

x=df.iloc[:,:-1]   #all but last coloumn
x=df.iloc[:.-1]  #only last coloumn

numpuy

mean=np.mean(data)
st_dev=np.std(data)
median=np.median(data)
var=np.var(data)