python Flashcards
Find the last char in str
str[-1]
Find the second to last char in str
str[-2]
Find the whole string apart for the last two chars
str[ :-2]
Find the whole string apart for the first two chars
str[ 2: ]
iterate over str
for i in range(len(str)):
Find the end of [ ] of numbers then check ==9
end = len(nums)
for i in range (end):
if nums [ i ] == 9:
return True
Look for one number in an array of numbers
for number in nums:
a function with if and then else if and else to find your fashion level.
def date_fashion(you, date):
if you < 3 or date < 3:
return 0
elif you >= 8 or date >=8 :
return 2
else:
return 1Find the position of w
astring = “Hello world!”
print(astring.index(‘w’))
Put a string in reverse
print(astring[::-1])
String to upper then lower
print(astring.upper())
print(astring.lower())
The environment
consists of Python standard libraries and pre-installed packages
What does a package always have
__init__.py
What error do you get if the package is missing in your system?
ModuleNotFoundError
List
[1,2,3,4]
List properties
- They are ordered.
- They can contain any objects.
- Their size can be varied.
- They are nestable that is
they can contain other lists
as elements. - Their elements can be accessed
by index. - They are mutable that is
you can perform various functions
on it.
The keys of the dictionary are
immutable , unordered and Unique
Tuples
immutable list.
heterogeneous sequence of elements,
impossible to append, edit or remove any individual elements within a tuple.
while Panel data
are observations over time, of the same characteristic for multiple entities
Using the map function
anw = list(map(lambda x,y: x+y , list_1, list_3 ))
An array ?
Array is it a list of related data types, [ [] [] [] ], used to multidimensional array
Numpy arrays advantage ?
Numpy arrays are helpful due to vectorisation and the ability to broadcast one row against another row, they can only work Arrays.
one dimensional arrays’ are know as ?
‘vectors’. ‘Scalars
NumPy array
np.array( [ [] [] [] ] )
print (A2[2:4:1,2:7:2]) what will happen
Brings row 2 & every digit
and brings columb 2, 6 every other digit
Make a few lists of the same type in np
V2 = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
to make up a 50 item array and then reshape it to 10 rows by 5 columns ?
np.linspace(1,50).reshape(10,5)
Example of a tuple ?
Tuples have no append or extend method.
Elements cannot be removed from a tuple.
(2,3,4,5,6)
Change a row to a column
B = np.array([1, 2, 3])
output = B[ : , np.newaxis ]
print (output)
Used to delaet column
DataFrame.drop()
Which of the following functions is used for indexing within a dataframe?
DataFrame.iloc()
What does the series constructor look like
pd.Series([10, 20, 30, 40, 50, 60])
how to times Series data with tan
My_Series = (S3+S4).apply(np.tan)
DataFrame drop index rows
shop = shop.drop(shop.index[0 : 247])
shop
out of dataframe slice first 3 lines and first 3 columns
df.iloc[:2,:3]
group items ?
df.groupby( [‘Market’, ‘Sector’] ).groups
Loop
for i in range(len(shop)):
if shop.iloc[i][“Close”] > 1420 :
print( shop.iloc[ i ] [ “Close” ] )
Work out percent change on row and column
shop.iloc[ 1 : 4 , 3 : 4 ].pct_change()
To create event driven trading
schedule strategy
stream data
place orders
How do you change from the ‘base’ environment to an environment named ‘quantra_py’?
. conda activate quantra_py
VWAP
Volume weighted average price
FInd help in Pythong
Help( ) or ? after item
print statement
print ( “ {0:1s} “ has {1 : .1 f} outstanding “ . format( name, number ) )
magic commands
%lmagic , %magic , %who
add list item
list . append (“jonathon”)
list . insert (0,”jon” )
ballian check in list
print ( “jon” in family_list)
True
Index using np
np.argmax(list)
Looking for items in a list > 200
np.where( (list >200) & (list <300) )
What is a set
set = {‘a’,’a’,’u’}
a = a + 1`
a += 1
how to get the current working dir
import os
os.getcwd()
FInd if about data frame
df.info() df.describe()
read in csv set index
data3 = pd.read_csv( “ shop.csv”, index_col=”Date”, parse_dates=True, dayfirst=True)
make a data frame
close_open = df [ [ ‘Close’, ‘Open’] ]
to delete column
del df[‘New’]
sum a whole column
df[‘Volume’].sum()
df open <1250
df [ ( df[‘Open’] < 1250 ) ]
selecting rows
df.iloc[ [ 4, 8, 20] ]
Selecting two rows and two columns
df.iloc[[4, 5], [0,3]]
closing change for day before
df[‘Close_to_Close’] = 100 * df[‘Close’] . pct_change()
Creating a new column called ‘Previous_Close’ by using the shift operator on Close column.
df[‘Previous_Close’] = df[ ‘Close’ ].shift(1)
df.head()
Creating a new column called ‘MA’ containing 5 day Moving Average of Close prices.
df[‘MA_5’] = df[ ‘Close’ ] . rolling (window=n).mean()
finding missing values
df.isnull().sum()
drop nana rows
df.dropna(axis=0).head()
fill forward
df.fillna( method=’ffill’ ).head()
function to find range then populate
def daily_range(x):
return x['Close']-x['Open']
df['daily_range'] = df.apply
( daily_range, axis=1 )supervised learning example
classification or regression learning
unsupervised learning
clustering
a set order doesn’t matter
{ , , ,}
what is the difference in continue and break
continue is more lie skip
break
Mak 1 col all 0
df.iloc { 10:50 , 6 } = 0
find out directories in package
dir( read_csv )
show all variables in memory
%%whos
work out average over 5 days
df{new_col} = df[ ‘close” ].rolling(5).mean()
Sum of all null values
df . isnull() . sum()
Change to datatime
df = pd.to_datetime( df.index )
logical condition
con = (df1_msft[“p”] > 0)
df1_msft.loc[ con ]
Make the first col an index
index_col= [ 0 ] , parse_dates=True
Help on function
yf . download?
today’s date
today = datetime.date.today()
to get all instance attributes
__dict__
family tree
__mro__
pass attributes and methods to child class
cls pass
Inherit methods from parents
Use supper
parent1.__init__(self,age) / parent2.__init__(self,age)
Find to days date
pd.datetime.now()
Date in the past
pd.Timedelta(days=30)
make a df
list1 =[ 1,2,3,4,5,6,7]
test_df = pd.DataFrame( { “name” : list1 } )
add to list
stock_list.extend( )
How to plot scatter
DataFrame.plot.scatter(test_df.plot.scatter( x=”name”,y=”name”,c=”red”, figsize=(12,5) ) )
work out log returns for stock
stock_df[stock] = np.log( [‘Close’] / stock_df[‘Open’] )
