markdown
stringlengths
0
1.02M
code
stringlengths
0
832k
output
stringlengths
0
1.02M
license
stringlengths
3
36
path
stringlengths
6
265
repo_name
stringlengths
6
127
1. Who Scored More Goals Per Year?
plt.suptitle('Goals Per Year For Fvery Player', fontsize = 15 ) ronaldo_df.groupby(['year'])['player'].count().plot(kind='bar',color='brown',label='Ronaldo',figsize=(15,5),width = 0.4) messi_df.groupby(['year'])['player'].count().plot(kind='bar',color='#14D622',label='Messi',figsize=(15,5),width = 0.2) plt.legend() plt...
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
From this comparison we can understand: 1. 2012 was a dramatic year in Messi and Ronaldo's career, it was the best year for Messi career's and the worst year for Ronaldo career's. 2. Until 2012 Messi scored more goals than Ronaldo, and from 2013 Ronaldo scores more goals. 3. Ronaldo is more stable than Messi across...
messi_type = df[df['player'] == 'Messi'].type.value_counts() ronaldo_type = df[df['player'] == 'Ronaldo'].type.value_counts() df_types = pd.DataFrame({ 'Messi' : messi_type , 'Ronaldo' : ronaldo_type}) df_types.T print("This graph show the number of goals in percent of the total goals for each ...
This graph show the number of goals in percent of the total goals for each player
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
From this comparison we can understand: 1. Messi scored 60% of his goals in his dominant foot, compared to Ronaldo who scored a little less than 40% in his dominant foot. 2. Ronaldo scored more goals on penalties and header. 3. Messi and Ronaldo scored the same percents with free kicks. 3. Who Scored More Solo Goa...
df['Solo_messi'] = ((df['assisted'] == 'Solo') & (df['player'] == 'Messi')) df['Solo_ronaldo']= ((df['assisted'] == 'Solo') & (df['player'] == 'Ronaldo')) Solo_ronaldo = (df['Solo_ronaldo'] == True).sum() Solo_messi = (df['Solo_messi'] == True).sum() nonSolo_messi = (len(df['Solo_messi']) - Solo_messi) nonSolo_ronaldo...
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
From this comparison we can understand: Ronaldo scored a few more solo goals than Messi. 4. Who Scored More Goals Per Possition?
df['pos']=df.pos.str.strip() messi_pos = df[df['player'] == 'Messi'].pos.value_counts() ronaldo_pos = df[df['player'] == 'Ronaldo'].pos.value_counts() df_pos = pd.DataFrame({ 'Messi' : messi_pos , 'Ronaldo' : ronaldo_pos}) df_pos = df_pos/df_pos.sum() df_pos.plot.bar(title = 'player favorite position', col...
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
From this comparison we can understand: 1. The best position for Ronaldo is LW and for Messi is RW.2. Messi better than Ronaldo in CF possition.3. Ronaldo better than Messi on the "weak side".4. Messi scored from more possionss than Ronaldo, but in low percents. 5. Who Was The Best Scorer In "LaLiga" (the spanish l...
plt.plot() ronaldo_df.loc[ronaldo_df.comp=='LaLiga'].groupby(['year'])['player'].count().plot(kind='line',color = 'brown',label='Ronaldo',figsize=(20,8)) messi_df.loc[(messi_df.comp=='LaLiga')].groupby(['year'])['player'].count().plot(kind='line',color = '#14D622',label='Messi',figsize=(20,6)) plt.title('Goals Scored I...
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
Average goals per player per season:
df.loc[df.comp=='LaLiga'].groupby(['player','year'])[['year']].count().groupby(["player"]).mean().style.set_caption("Average goals per player per season")
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
From this comparison we can understand: 1. Ronaldo is more stable than Messi across the years.2. In the period from 2009 to 2018, Messi was the best scorer for 6 seasons (2009, 2010, 2012, 2016, 2017, 2018) compared to only 4 seasons for Ronaldo (2011, 2013, 2014, 2015).3. Ronaldo scores an average of 2 goals more t...
plt.subplot(1,2,1) df.loc[(df['comp']=='Champions League') & (df['round'] == 'Group Stage')].groupby(['player'])['date'].count().plot(kind='bar',width = 0.2, color = ( "#14D622","brown")) plt.title('Goals Scored in "UEFA Champions League" Group stages',fontsize = 20) plt.xlabel('Players',fontsize = 14) plt.ylabel('Numb...
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
Goals Scored in UEFA Champions League Knockout stages:
df.loc[(df['comp']=='Champions League') & (df['round'] == 'Group Stage')].groupby(['player','comp'])[['date']].count()
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
Goals Scored in "UEFA Champions League" Group stages:
df.loc[(df['comp']=='Champions League') & (df['round'] != 'Group Stage')].groupby(['player','comp'])[['date']].count()
_____no_output_____
CC0-1.0
projects/projects_summer_2021/15_football.ipynb
nlihin/my-binder
Introduction to Python 3 in jupyternotebookThis is an introduction to the Python programming language for participants on the Programming with openBIM course as arranged by [BIMFag](https://bimfag.no/). It is based on [learnpythons basic course](https://www.learnpython.org/). It is not a complete Python course, so for...
### Change this cell from a Code cell to a Markdown cell, by using the roll down menu above.
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Let the program tell something to a userThe print statement is a powerfull directive in python and it enables a program to print out a result, eg. "Hello, world". Which prints a data type called a string to the screen.
# Here we print the string "Hello, world!" to the screen. print("Hello, world!")
Hello, world!
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
We will use the print statement several times in this course together with "string formatting". A typpical way of formatting a string is by appending another string to it, like "Hello, "+"world".
# Here we append the two strings "Hello, " and "world!" to the screen. print("Hello,"+"world!")
Hello,world!
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
As seen above now we didn't get a space between the two strings. A space could either be part of e.g. the first string, or be a separate string in itself like " ". Lets append that too.
# Here we append the three strings "Hello,", " " and "world!" print("Hello,"+" "+"world!")
Hello, world!
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Let the program ask a user for input and show the resultSince we want to say hello to you too, you could use the "input()" function to get user intput and store that in a variable, and print to screen.
# Here we get the name from the user and store it in a variable called name, append that to another vairable as we print. x = input("Write your name here: ") y = "Hello, " print(y+x+"!")
Write your name here: Ingvild Hello, Ingvild!
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Variables and types Python is completely object oriented, and not "statically typed". In other programming languages, that is statically typed, you need to declare the variable and what type of data it holds, before using it. In Python you do not need to declare variables before using them, or declare their type. This...
# This is how you define an integer myInt = 9 print(myInt) # To define a floating point number you could use one of the following approaches myFloat1 = 9.0 print(myFloat1) myFloat2 = float(9) print(myFloat2) myFloat3 = 9. print(myFloat3)
9.0 9.0 9.0
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
One could aslo "cast" an integer into a float, like we do in the myFloat2=float(9). * An integer x can be cast into float by using float(x)* A float y can be cast into an integer by using int(y)* An integer x and a float y can both be cast into a string by respectively str(x) and str(y).
# Here we first define an integer, and cast it into a float myInt = 9 myFloat = float(myInt) print("My Integer: %s, and my Float: %s" % (myInt,myFloat)) # Here we cast myInt and myFloat (from above) into strings myIntegerString = str(myInt) print(myIntegerString) myFloatString= str(myFloat) print(myFloatString)
9 9.0
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Strings Strings can be define by using single (') or double quoates("). The difference between the two is that using double quotes makes it easy to include apostrophes (whereas these would terminate the string if using single quotes)
myString = "Don't worry about apostrophes" print(myString) escapedString = 'Don\'t worry about apostrophes' print(escapedString) #Strings are just an array of characters myString = "abcdefgh" print(myString) third = myString[2] print(third) last = myString[-1] print(last)
abcdefgh c h
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
MiniExcercise:Fetch the character "d" from myString and print it to screen
# Write your code here
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 1 The target of this exercise is to create a string, an integer, and a floating point number. The string should be named mystring and should contain the word "hello". The floating point number should be named myfloat and should contain the number 10.0, and the integer should be named myint and should contain ...
# change this code mystring = "change this" myfloat = "change this" myint = "change this" # testing code if mystring == "hello": print("String: %s" % mystring) if isinstance(myfloat, float) and myfloat == 10.0: print("Float: %f" % myfloat) if isinstance(myint, int) and myint == 20: print("Integer: %d" % my...
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 2Run the code below first without changing anything, supplying a number to it. Then, use a cast to correct it so that it outputs correct math on numbers.
# Here we take and input of a number, and then multiply it 5 times and then prints the result. inputVariable = input("Provide a number: ") x = inputVariable * 5 print(x)
Provide a number: 6 66666
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Naming conventions:Different programming languages have different "best practice" naming for their naming of variables, funtions and classes. In python, the naming conventions are described at: [PEP8](https://www.python.org/dev/peps/pep-0008/naming-conventions)
# Examples of naming conventions CONSTANTS_ARE_ALL_UPPERCASE = "my constant string" variables_should_be_all_lowercase_separated_with_underscores = 1 def functions_should_do_the_same(): var1 = 4 retrun var1
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Basic String formattingAs seen in the testing code: ```Python testing codeif mystring == "hello": print("String: %s" % mystring)if isinstance(myfloat, float) and myfloat == 10.0: print("Float: %f" % myfloat)if isinstance(myint, int) and myint == 20: print("Integer: %d" % myint)```We use some string formatting...
# Execute this to check multiple variables inserted into a string print("string %s, number %d and floatingpoint %f without formatting, or formatted %0.5f" % ("hello",10,10.0, 10.0))
string hello, number 10 and floatingpoint 10.000000 without formatting, or formatted 10.00000
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
MiniExercise 2B:Try doing float printing with only two decimals
# Execute this to check multiple variables inserted into a string print("Print the float %f" % (10.0))
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
String functionsIn addition to this there are several other string formatting options that you could do. Strings are objects that have several predefined functions. Feel free to add some code cells below here with some of the example code snippets from [w3schools on strings](https://www.w3schools.com/python/python_str...
# Example 1 -- eg. convert a string to lower letters a = "HElLo!" print(a.lower()) # covert to captial letters print(a.upper()) # Example 2 -- eg. split a string into two. a = "Hello, world!" #split at a specific character var = a.split(",") print(var) #split whitespace at beginning or end print(var[1].split())
['Hello', ' world!'] ['world!']
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 3Build a script in the cell below that: 1) Takes in a string that you input2) Make sure it is a string3) Store it in a variable 4) Print out the variable5) Then grab only the first and last characters in the name6) Store it in a new variable7) Print out the new variablesee [w3schools on strings](https://www.w...
### Exercise 3 answer here:
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Collection types (Arrays)There are four collection data types in the Python programming language:* List is a collection which is ordered and changeable. Allows duplicate members.* Tuple is a collection which is ordered and unchangeable. Allows duplicate members.* Set is a collection which is unordered and unindexed. N...
myList = ["apple", 1, "cherry", 10] print(myList)
['apple', 1, 'cherry', 10]
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
You access the list items by referring to the index number. Remember that the index starts at 0.
myList = [1,2,3,4,5] print(myList[1])
2
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Lists are changableLists are changable, so you could eg. add or remove from it.
# Adding to a list my append() myList = [1,2,3] myList.append(3) # adds the value 3 to the list print(myList) # Removing from a list by remove() myList.remove(2) # removes the value 2 from the list print(myList) # Removing from a list by pop() myList.pop() # removes the last item in the list print(myList) myList.pop(...
[1, 2, 3, 3] [1, 3, 3] [1, 3] [3]
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
TupleA tuple is a collection which is ordered and **unchangeable**. In Python tuples are written with parenthesis. When working with IFC and the ifcopenshell, we often work with tuples.
myTuple = ("apple", 1, "cherry", 10) print(myTuple)
('apple', 1, 'cherry', 10)
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
You access the tuple items, similar as for lists, by referring to the index number. Remember that the index starts at 0.
myTuple = (1,2,3,4,5) print(myTuple[1])
2
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Tuples are unchangableTuples are unchangable, but you could cast it into a list and work with it as a list.
myTuple = (1,2,3,4,5) myList = list(myTuple) # Cast tuple into a list myList.pop(2) myTuple = tuple(myList) # Cast list into a tuple print(myTuple)
(1, 2, 4, 5)
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Check the number of elements in list or tupleIt is often needed to check how many elements that are in a list or a tuple.
# Check how many elements there are in myList (defined above) print(len(myList)) # Check how many elements there are in myTuple (defined above) print(len(myTuple))
3 3
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 4 - how many IfcWall elements are in Grethes Hus model? In the excercise below we want you to use what you have learned above to find how many elements of type IfcWall are in the Grethes Hus Modell. We have provided some code to get the file and to query out all elements of type IfcWall and stored it in a var...
# We import ifcopenshell and open the Grethes Hus Model and store it in a variable called file. import ifcopenshell file = ifcopenshell.open("models/Grethes-hus-bok-2.ifc") #We query the file for its number of IfcWall type elements walls = file.by_type("IfcWall") """ToDo: Change the zero below to list out the numbe...
0
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Conditions and If statements > Python supports the usual logical conditions from mathematics:> * Equals: a == b> * Not Equals: a != b> * Less than: a < b> * Less than or equal to: a <= b> * Greater than: a > b> * Greater than or equal to: a >= b>> These conditions can be used in several ways, most commonly in "if stat...
## Here we refer back to the variable "numberOfWalls" as defined above, and check if its greated than 0. if numberOfWalls > 0: print("You probably did something right above.")
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
elif and elseWe could also add to the If statement, by using an **Elif** statement. This will allow you to add another condition to check. At the end, we could also do an **Else** statement. This will be executed if no of the previous conditions where met. The syntax for this is: ```Python if : do something...
if numberOfWalls > 0: print("You probably did something right above.") elif numberOfWalls == 0: print("You did probably do something wrong above") else: print("How could this happen...?")
You did probably do something wrong above
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Check if an element exist in a list or tupleOften times one would like to check if a particular value is in a list or a tuple. This is easy with python.
# Check if a value is in a list myList = ["BIMFag", "banana", "BIM"] if "BIMFag" in myList: print("I found BIMFag in myList!") myTuple = ["BIMFag", "banana", "BIM"] if "BIMFag" in myTuple: print("I found BIMFag in myTuple!")
I found BIMFag in myList! I found BIMFag in myTuple!
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
While LoopsConditions and if statements are usefull also in while loops. While loops are executing as long as a condition is met. The syntax for while loops are: ```Python x = 1 while x<10: Checks if x is less than 10 do something eg. print(x) since x is bigger than zero this will print for as long ...
# Lets see it in action. Print out x as long as its less than the number of walls x = 1 while x < numberOfWalls: print(x) x = x +1 else: print("x is: %s and numberOfWalls is: %s"%(x,numberOfWalls))
x is: 1 and numberOfWalls is: 0
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Loops, break and elseNotice that the loop exits when the condition is no longer met. Notice also that loops could also have an else statement, similar to if statements. If the loop condition are no longer met, then the else part is executed. If we want to exit the while loop early one could use the special **break** s...
x = 1 ## Loop while True --> True is always True... while True: print(x) # check if x is bigger or eaqual to 10 if x >= 10: break # if condition met, break out of the while loop. x = x+1 # if the condition of the if statement was false, this get executed in the while loop.
1 2 3 4 5 6 7 8 9 10
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
For LoopsA *for* loop is used for iterating over a sequence (that is either a list, a tuple, a dictionary, a set, or a string).The syntax for for loops in python is: ```Pythonfor element in listOfElements: Do something with each element eg. print it out. print(element)```Notice the **indentation** here as well....
myTuple = (1,2,3,4,5,6) for element in myTuple: print(element) for i in range(6): print(myTuple[i])
1 2 3 4 5 6
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Iterating over a number of IfcWindow objects and access attributesBelow we'll use a for loop to loop through all IfcWindow elements and print out some of it's direct attributes. See the IFC data dictinary for IFC 2.3 [here](https://standards.buildingsmart.org/IFC/RELEASE/IFC2x3/FINAL/HTML/). The main access page for t...
# We import ifcopenshell and open the Grethes Hus Model and store it in a variable called file. import ifcopenshell file = ifcopenshell.open("models/Grethes-hus-bok-2.ifc") #We query the file for its number of IfcWindow type elements windows = file.by_type("IfcWindow") """By uncommenting (removing the '#') the line...
The window elements name is: M_Fixed:_1400x2200mm:348960 The window elements name is: M_Fixed:_1400x2200mm:350239 The window elements name is: M_Fixed:_1400x2200mm:350479 The window elements name is: M_Fixed:_1400x2200mm:350484 The window elements name is: M_Fixed:_1400x2200mm:350489 The window elements name is: M_Fixe...
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Functions Functions are a useful way to divide your code into blocks of code, that only runs when they are called. It is also a good way to share a small snippet of code with others. A function is defined by the special ```def```statement. Like the example below. Defining a function
def myDefinedFunction(): print("This is a print statement within my function")
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Notice that the function uses the same indendation style as for If statements and loops as walked through above. A function could contain all the concepts we have walked through above as well. Notice however that noting happended above. That is why we havent called the function. It is however stored, so that we can cal...
myDefinedFunction()
This is a print statement within my function
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Input variables to functionsWe could also build functions that takes in input variables that could be used in the function. Lets define a function that takes in a string variable and prints it out.
# A function definition that takes in a parameter called "string" def myDefinedFunction2(string): print("My variable is %s"%string) # A call to myDefinedFunction2 with the string variable "Hello again!" myDefinedFunction2("Hello again!") # A new call to myDefinedFunction2 with the string "and again and again" myDe...
My variable is and again and again
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Input variables and assumed typeWhat would happen if we call myDefinedFunction2 with a number, instead of a string? Try it below.
string = 10 """Try calling myDefinedFunction2 with the variable define above"""
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 5 - ensure that a function works as intendedYou can pass as many variables to a function as you'd like, just separate them with a comma like we do with var1 and var 2 below. They could be whatever object you'd like. Below we want to create a function that takes in two variables, var1 and var2. If they are bot...
def myFunc3(var1,var2): """Fill in code to comlete the challenge""" """Don't change the code below""" myFunc3("Hello, ","World!") myFunc3("hello",4)
Hello, World! ['hello', 4]
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Lists and tuples as input variablesWe could also add lists and tuples as input variables to functions.
def myFunc4(myList): for elem in myList: print(elem) # Define a list aList = [1,2,3] # Send it as input variable to myFunc4 myFunc4(aList) # Define a tuple aTuple = (4,5,6) # Send it as input variable to myFunc4 myFunc4(aTuple)
1 2 3 4 5 6
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
What will happend if you pass a string instead of a list?Try below to call the function above passing in a string value like eg. ```myFunc4("your name")```What do you think will happen?
### Do the test here ###
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Have functions return valuesA good way to use functions is to have them perform som logic on input variables and have them return the result. For this you use the ```return``` statement.
# defines a function that takes in a list of items as parameter, # loops through it and adds all integers together, and then returns the resulting value def myFunc5(myList): tmp = 0 # Defined outside the for loop to be able to return it outside the scope of the loop for elem in myList: if isinstance(el...
54
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 6 - Names and Descriptions of Window objects. Many times one would get an ifc model where one want to change or restructure information in the file. That is tediouse work on big models, so one may want to script it. In addition it would be a good code block to have, regardless of object type. In IFC the name ...
### Add your code here ###
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Classes and ObjectsWe have already worked alot with objects, some window objects, some string objects, and even some list objects. So, what is objects and where do they come from? Objects are an encapsulation of variables and functions into a single entity. Objects get their variables and functions from classes. Class...
## Defines a Window Class class MyWindowClass: nameVariable = "Window" def namePrint(self): print("My Name is "+self.nameVariable)
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Creating objects from class definitionsWhen you have defined a class, you could create several objects based on it.
## Defines two different objects based on the MyWindowClass window1 = MyWindowClass() window2 = MyWindowClass()
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Accessing Object variables and functionsNow the window1 and window2 objects have both a nameVariable and namePrint function each. As we have seen above, we access its variables (attributes) by the "." operator.
# Get the nameVariable of window1 window1.nameVariable
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
MiniExercise: Do the same for window2 below
### Get the the name variable of window2
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 7 - Setting object variablesBoth windows have the same name. Use what you know from above to set the name of window1 and window2 to something proper.
### Solve Exercise 7 here:
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Exercise 8 - Accessing object functionsIn the MyWindowClass definition we have a function to print out the name. Show below how this function can be accessed on both window1 and window2.
### Solve Exercise 8 here:
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Modules and PackagesIn programming, a module is a piece of software that has a specific functionality. For example, when building a ping pong game, one module could be responsible for the game logic, and another module could be responsible for drawing the game on the screen. Each module is a different file, which can ...
# Import the ifc_viewer class of ifc_viewer module
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
Using Packages and Modules in another programAnother package we already have used is the **ifcopenshell**. Packages are namespaces which contain multiple packages and modules themselves. They are simply directories, but with a twist.Each package in Python is a directory which MUST contain a special file called __init_...
import ifcopenshell import ifcopenshell.geom from ifc_viewer import ifc_viewer # Storing the model in a file variable, and giving the path to the file as input. file = ifcopenshell.open("../models/Grethes_hus_bok_2.ifc") # Storing all windows of the file by ising the by_type function of the file class windows = fil...
_____no_output_____
MIT
Course 1/Notebooks/01_Basic_Introduction_to_Python_for_use_with_BIM.ipynb
nasirkuvvetli/intro-python-bim
04_2 Model_Stacking This notebook includes the data preparation and the developement of a Stacking model.Due to NDA agreements no data can be displayed. Data Preparation, Data Cleaning, and Preparation for Modelling is the same for all algorithms. To directly go to modelling click [here](modelling) --- Data preparati...
import pandas as pd import numpy as np from sklearn.metrics import mean_squared_error from sklearn.pipeline import make_pipeline from sklearn.preprocessing import MinMaxScaler from sklearn.model_selection import train_test_split import seaborn as sns import matplotlib.pyplot as plt import sys sys.path.append("..") ...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Create data frame with important features So that everyone is on track with the feature selection, we created another csv file to rate the importance and only use important features for training our models and further analysis. Only important features are used to train the model. In this case we use 17 features beside...
# read list with feature importance data_log = pd.read_csv('../data/Capstone_features_Features.csv') data_log.head() # create list of important features (feature importance < 3) list_imp_feat = list(data_log[data_log['ModelImportance'] < 3]['VarName']) len(list_imp_feat) df_model = df[list_imp_feat].copy() df_model.inf...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Fill and drop NaN Values for V.SLPOG.act.PRC and ME.SFCI.act.gPkWh contain missing values. The EDA showed that these are mainly caused during harbour times when the main engine was not running. Therefore it makes sense to fill the missing values with 0.
df_model['V.SLPOG.act.PRC'].fillna(0,inplace=True) df_model['ME.SFCI.act.gPkWh'].fillna(0,inplace=True) df_model['A.SOG.next.kn'] = (df_model['V.SOG.act.kn'].shift(-1) - df_model['V.SOG.act.kn']) df_model['A.SOG.next.kn'].fillna(df_model['V.SOG.act.kn'], inplace=True) df_model['A.SOG.next.kn'].describe()
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
The remaining rows with missing values are dropped.
df_model.dropna(inplace=True) df_model.info() plt.figure(figsize = (30,28)) sns.heatmap(df_model.corr(), annot = True, cmap = 'RdYlGn')
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Define target
X = df_model.drop(['ME.FMS.act.tPh'], axis = 1) y = df_model['ME.FMS.act.tPh'] X.rename(columns={'passage_type_Europe<13.5kn': 'passage_type_Europe_smaller_13.5kn', 'passage_type_Europe>13.5kn': 'passage_type_Europe_greater_13.5kn',\ 'passage_type_SouthAmerica<13.5kn': 'passage_type_SouthAmerica_smaller_13.5kn', 'p...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Train Test Split Due to the high amount of data, a split into 10% test data and 90% train data is chosen. The random state is set to 42 to have comparable results for diffent models. To account for the imbalance in the distribution of passage types the stratify parameter is used for this feature. This results in appro...
X_train, X_test, y_train, y_test = train_test_split(X, y, stratify = X['passage_type'], test_size = 0.1, random_state = 42)
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Create dummy values for passage type As passage_type is the only object type, get_dummies will only create dummies for passage_type.
X_train = pd.get_dummies(X_train, drop_first=True) X_test = pd.get_dummies(X_test, drop_first=True)
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Set MLFlow connection MLFlow is used to track and compare different models and model settings.
runmlflow = False # setting the MLFlow connection and experiment if runmlflow == True: mlflow.set_tracking_uri(TRACKING_URI) mlflow.set_experiment(EXPERIMENT_NAME) mlflow.start_run(run_name='Stacking (Poly, RF_Hyper)') # CHANGE! run = mlflow.active_run()
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
--- Modelling
RSEED = 42
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
For all models in this project a MinMaxScaler is applied. For this model a random forrest is used. The hyperparameter are selected based on grid search and offer a reasonable balance between optimal results and overfitting. These settings are used in a pipeline. Pipeline
estimators = [ ('rfh', make_pipeline(MinMaxScaler(), RandomForestRegressor(criterion= 'squared_error', max_depth= 40, max_features= 'auto', max_leaf_nodes= 7000, ...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Fit and predict
reg.fit(X_train, y_train) y_pred = reg.predict(X_test) y_pred_train = reg.predict(X_train)
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
--- Analysis Errors and residuals The root mean squared error (RMSE) is used to evaluate the model.
y_pred2 = y_pred.copy() y_pred2_train = y_pred_train.copy() y_pred2[y_pred2 < 0.013509] = 0 y_pred2_train[y_pred2_train < 0.013509] = 0 #0.013509 print('RMSE train: ', mean_squared_error(y_train, y_pred2_train, squared= False)) rmse_train = mean_squared_error(y_train, y_pred2_train, squared= False) print('RMSE test: ...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Plotting actual values against predicted shows that the points are close to the optimal diagonale. However, this plot and the yellowbrick residual plot show some dificulties the model has when predicting low target values.
fig=plt.figure(figsize=(6, 6)) plt.axline([1, 1], [2, 2],color='lightgrey') plt.scatter(y_train, y_pred2_train, color ='#33424F') plt.scatter(y_test, y_pred2, color = '#FF6600') #plt.xticks(np.arange(0,501,100)); #plt.yticks(np.arange(0,501,100)); plt.xlabel("ME.FMS.act.tPh actual"); plt.ylabel("ME.FMS.act.tPh predicte...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
--- Write to MLFlow
#seting parameters that should be logged on MLFlow #these parameters were used in feature engineering (inputing missing values) #or parameters of the model (fit_intercept for Linear Regression model) params = { "features drop": 'EntryDate,Date_daily, Type_daily, TI.LOC.act.ts, WEA.WDR.act.deg, WEA.WSR.act.mPs, WE...
_____no_output_____
MIT
notebooks/04_2_Model_Stacking.ipynb
SuRreal1000/capstone_know_your_ship
Theta Sketch Examples Basic Sketch Usage
from datasketches import theta_sketch, update_theta_sketch, compact_theta_sketch from datasketches import theta_union, theta_intersection, theta_a_not_b
_____no_output_____
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
To start, we'll create a sketch with 1 million points in order to demonstrate basic sketch operations.
n = 1000000 k = 12 sk1 = update_theta_sketch(k) for i in range(0, n): sk1.update(i) print(sk1)
### Update Theta sketch summary: lg nominal size : 12 lg current size : 13 num retained keys : 6560 resize factor : 8 sampling probability : 1 seed hash : 37836 ordered? : false theta (fraction) : 0.00654224 theta (raw 64-bit) : 603415087386602...
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
The summary contains most data fo interest, but we can also query for specific information. And in this case, since we know the exact number of distinct items presented ot the sketch, we can look at the estimate, upper, and lower bounds as a percentage of the exact value.
print("Upper bound (1 std. dev) as % of true value:\t", round(100*sk1.get_upper_bound(1) / n, 4)) print("Sketch estimate as % of true value:\t\t", round(100*sk1.get_estimate() / n, 4)) print("Lower bound (1 std. dev) as % of true value:\t", round(100*sk1.get_lower_bound(1) / n, 4))
Upper bound (1 std. dev) as % of true value: 101.5208 Sketch estimate as % of true value: 100.2715 Lower bound (1 std. dev) as % of true value: 99.0374
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
We can serialize and reconstruct the sketch. If we compact the sketch prior to serialization, we can still query the rebuilt sketch but cannot update it further.
sk1_bytes = sk1.compact().serialize() len(sk1_bytes) new_sk1 = theta_sketch.deserialize(sk1_bytes) print("Estimate: \t\t", new_sk1.get_estimate()) print("Estimation mode: \t", new_sk1.is_estimation_mode())
Estimate: 1002714.745231455 Estimation mode: True
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
Sketch Unions Theta Sketch unions make use of a separate union object. The union will accept input sketches with different values of $k$.For this example, we will create a sketch with distinct values that partially overlap those in `sk1`.
offset = int(3 * n / 4) sk2 = update_theta_sketch(k+1) for i in range(0, n): sk2.update(i + offset) print(sk2)
### Update Theta sketch summary: lg nominal size : 13 lg current size : 14 num retained keys : 12488 resize factor : 8 sampling probability : 1 seed hash : 37836 ordered? : false theta (fraction) : 0.0123336 theta (raw 64-bit) : 113757656857900...
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
We can now feed the sketches into the union. As constructed, the exact number of unique values presented to the two sketches is $\frac{7}{4}n$.
union = theta_union(k) union.update(sk1) union.update(sk2) result = union.get_result() print("Union estimate as % of true value: ", round(100*result.get_estimate()/(1.75*n), 4))
Union estimate as % of true value: 99.6787
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
Sketch Intersections Beyond unions, theta sketches also support intersctions through the use of an intersection object. These set intersections can have vastly superior error bounds than the classic inclusion-exclusion rule used with sketches like HLL.
intersection = theta_intersection() intersection.update(sk1) intersection.update(sk2) print("Has result: ", intersection.has_result()) result = intersection.get_result() print(result)
Has result: True ### Compact Theta sketch summary: num retained keys : 1668 seed hash : 37836 ordered? : true theta (fraction) : 0.00654224 theta (raw 64-bit) : 60341508738660257 estimation mode? : true estimate : 254959 lower bound 95% conf : 242...
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
In this case, we expect the sets to have an overlap of $\frac{1}{4}n$.
print("Intersection estimate as % of true value: ", round(100*result.get_estimate()/(0.25*n), 4))
Intersection estimate as % of true value: 101.9834
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
Set Subtraction (A-not-B) Finally, we have the set subtraction operation. Unlike `theta_union` and `theta_intersection`, `theta_a_not_b` always takes as input 2 sketches at a time, namely $a$ and $b$, and directly returns the result as a sketch.
anb = theta_a_not_b() result = anb.compute(sk1, sk2) print(result)
### Compact Theta sketch summary: num retained keys : 4892 seed hash : 37836 ordered? : true theta (fraction) : 0.00654224 theta (raw 64-bit) : 60341508738660257 estimation mode? : true estimate : 747756 lower bound 95% conf : 726670 upper bound...
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
By using the same two sketches as before, the expected result here is $\frac{3}{4}n$.
print("A-not-B estimate as % of true value: ", round(100*result.get_estimate()/(0.75*n), 4))
A-not-B estimate as % of true value: 99.7008
BSL-1.0
python/jupyter/ThetaSketchNotebook.ipynb
tdoehmen/datasketches-cpp
SVM
X_train, X_test, Y_train, Y_test = train_test_split(X_train_transform, sentiment, test_size=0.3) clf = SVC().fit(X_train, Y_train) predicted = clf.predict(X_test) print(classification_report(Y_test, predicted)) kf = KFold(n_splits=10) clf = SVC() lista = [] for train_index, test_index in kf.split(X_train_transform): ...
0.3835125448028674 0.45878136200716846 0.4121863799283154 0.43727598566308246 0.4659498207885305 0.44802867383512546 0.44086021505376344 0.4748201438848921 0.44964028776978415 0.48201438848920863 Média: 0.44530698022227383 Std: 0.028021649950196455
MIT
Tash-PT/RNN_tash.ipynb
letisousa/SentimentAnalysisUpdates
RL
X_train, X_test, Y_train, Y_test = train_test_split(X_train_transform, sentiment, test_size=0.3) clf = LogisticRegression(max_iter=1000).fit(X_train, Y_train) predicted = clf.predict(X_test) print(classification_report(Y_test, predicted)) kf = KFold(n_splits=10) clf = LogisticRegression(max_iter=1000) lista = [] for ...
_____no_output_____
MIT
Tash-PT/RNN_tash.ipynb
letisousa/SentimentAnalysisUpdates
CONV1D
model = tf.keras.Sequential([ vectorize_layer, tf.keras.layers.Embedding( input_dim=len(vectorize_layer.get_vocabulary()), output_dim=64, mask_zero=True), tf.keras.layers.Conv1D(32,6, activation='relu'), tf.keras.layers.MaxPooling1D(2), tf.keras.layers.Flatten(), ...
Fold: 0 Ultimo valor acc: 0.41577062010765076 Fold: 1 Ultimo valor acc: 0.379928320646286 Fold: 2 Ultimo valor acc: 0.40143370628356934 Fold: 3 Ultimo valor acc: 0.3476702570915222 Fold: 4 Ultimo valor acc: 0.4444444477558136 Fold: 5 Ultimo valor acc: 0.379928320646286 Fold: 6 Ultimo valor acc: 0.37992832...
MIT
Tash-PT/RNN_tash.ipynb
letisousa/SentimentAnalysisUpdates
BDR
model = tf.keras.Sequential([ vectorize_layer, tf.keras.layers.Embedding( input_dim=len(vectorize_layer.get_vocabulary()), output_dim=64,mask_zero=True), tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(50)), tf.keras.layers.Dense(16, activation='relu'), tf.keras.layers.D...
_____no_output_____
MIT
Tash-PT/RNN_tash.ipynb
letisousa/SentimentAnalysisUpdates
LSTM
model = tf.keras.Sequential([ vectorize_layer, tf.keras.layers.Embedding( input_dim=len(vectorize_layer.get_vocabulary()), output_dim=64,mask_zero=True), tf.keras.layers.LSTM(50, activation='relu' ,return_sequences=True), tf.keras.layers.Dropout(0.3), tf.keras.layers.LSTM(1...
INFO:tensorflow:Reloading Oracle from existing project .\text_classifier\oracle.json INFO:tensorflow:Reloading Tuner from .\text_classifier\tuner0.json INFO:tensorflow:Oracle triggered exit Epoch 1/10 61/61 [==============================] - 6s 96ms/step - loss: 0.6423 - accuracy: 0.3564 - val_loss: 0.6409 - val_accura...
MIT
Tash-PT/RNN_tash.ipynb
letisousa/SentimentAnalysisUpdates
Training an Encrypted Neural NetworkIn this tutorial, we will walk through an example of how we can train a neural network with CrypTen. This is particularly relevant for the Feature Aggregation, Data Labeling and Data Augmentation use cases. We will focus on the usual two-party setting and show how we can train an ac...
import crypten import torch crypten.init() torch.set_num_threads(1) %run ./mnist_utils.py --option features --reduced 100 --binary
_____no_output_____
MIT
tutorials/Tutorial_7_Training_an_Encrypted_Neural_Network.ipynb
mh739025250/CrypTen
Next, we'll define the network architecture below, and then describe how to train it on encrypted data in the next section.
import torch.nn as nn import torch.nn.functional as F #Define an example network class ExampleNet(nn.Module): def __init__(self): super(ExampleNet, self).__init__() self.conv1 = nn.Conv2d(1, 16, kernel_size=5, padding=0) self.fc1 = nn.Linear(16 * 12 * 12, 100) self.fc2 = nn.Linear(1...
_____no_output_____
MIT
tutorials/Tutorial_7_Training_an_Encrypted_Neural_Network.ipynb
mh739025250/CrypTen
Encrypted TrainingAfter all the material we've covered in earlier tutorials, we only need to know a few additional items for encrypted training. We'll first discuss how the training loop in CrypTen differs from PyTorch. Then, we'll go through a complete example to illustrate training on encrypted data from end-to-end....
# Define source argument values for Alice and Bob ALICE = 0 BOB = 1 # Load Alice's data data_alice_enc = crypten.load_from_party('/tmp/alice_train.pth', src=ALICE) # We'll now set up the data for our small example below # For illustration purposes, we will create toy data # and encrypt all of it from source ALICE x_sm...
Epoch: 0 Loss: 0.3058 Epoch: 1 Loss: 0.2807
MIT
tutorials/Tutorial_7_Training_an_Encrypted_Neural_Network.ipynb
mh739025250/CrypTen
A Complete ExampleWe now put these pieces together for a complete example of training a network in a multi-party setting. As in Tutorial 3, we'll assume Alice has the rank 0 process, and Bob has the rank 1 process; so we'll load and encrypt Alice's data with `src=0`, and load and encrypt Bob's data with `src=1`. We'll...
import crypten.mpc as mpc import crypten.communicator as comm # Convert labels to one-hot encoding # Since labels are public in this use case, we will simply use them from loaded torch tensors labels = torch.load('/tmp/train_labels.pth') labels = labels.long() labels_one_hot = label_eye[labels] @mpc.run_multiprocess(...
torch.Size([100, 28, 20]) torch.Size([100, 28, 8])
MIT
tutorials/Tutorial_7_Training_an_Encrypted_Neural_Network.ipynb
mh739025250/CrypTen
We see that the average batch loss decreases across the epochs, as we expect during training.This completes our tutorial. Before exiting this tutorial, please clean up the files generated using the following code.
import os filenames = ['/tmp/alice_train.pth', '/tmp/bob_train.pth', '/tmp/alice_test.pth', '/tmp/bob_test.pth', '/tmp/train_labels.pth', '/tmp/test_labels.pth'] for fn in filenames: if os.path.exists(fn): os.remove(fn)
_____no_output_____
MIT
tutorials/Tutorial_7_Training_an_Encrypted_Neural_Network.ipynb
mh739025250/CrypTen
Line Continuation Add a backslash in the code below, so it is a one-line code. Observe the change in the result.
15 + 31\ - 26
_____no_output_____
MIT
Python basics practice/Python 3 (5)/Line Continuation - Exercise_Py3.ipynb
rachithh/data-science
Add a default legend to LayerIn this example, a default Legend with a title, description, and footer is added to the map.For more information, run `help(Layer)` or `help(Legend)`.
from cartoframes.auth import set_default_credentials from cartoframes.viz import Map, Layer, Legend set_default_credentials('cartoframes') Map( Layer( 'global_power_plants', legend=Legend( 'default', title='Global Power Plants', description='Power plant locations...
_____no_output_____
BSD-3-Clause
examples/layers/add_default_legend_layer.ipynb
jorisvandenbossche/cartoframes
Vertex pipelines**Learning Objectives:**Use components from `google_cloud_pipeline_components` to create a Vertex Pipeline which will 1. train a custom model on Vertex AI 1. create an endpoint to host the model 1. upload the trained model, and 1. deploy the uploaded model to the endpoint for serving OverviewThi...
!pip3 install --user google-cloud-pipeline-components==0.1.1 --upgrade
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
Restart the kernelAfter you install the additional packages, you need to restart the notebook kernel so it can find the packages. Import libraries and define constants
import os from datetime import datetime import kfp from google.cloud import aiplatform from google_cloud_pipeline_components import aiplatform as gcc_aip from kfp.v2 import compiler from kfp.v2.dsl import component from kfp.v2.google import experimental
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
Check the versions of the packages you installed. The KFP SDK version should be >=1.6.
print("KFP SDK version: {}".format(kfp.__version__))
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
Set your environment variablesNext, we'll set up our project variables, like GCP project ID, the bucket and region. Also, to avoid name collisions between resources created, we'll create a timestamp and append it onto the name of resources we create in this lab.
# Change below if necessary PROJECT = !gcloud config get-value project # noqa: E999 PROJECT = PROJECT[0] BUCKET = PROJECT REGION = "us-central1" TIMESTAMP = datetime.now().strftime("%Y%m%d%H%M%S") PIPELINE_ROOT = f"gs://{BUCKET}/pipeline_root" print(PIPELINE_ROOT)
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
We'll save pipeline artifacts in a directory called `pipeline_root` within our bucket. Validate access to your Cloud Storage bucket by examining its contents. It should be empty at this stage.
!gsutil ls -la gs://{BUCKET}/pipeline_root
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
Give your default service account storage bucket accessThis pipeline will read `.csv` files from Cloud storage for training and will write model checkpoints and artifacts to a specified bucket. So, we need to give our default service account `storage.objectAdmin` access. You can do this by running the command below in...
@component def training_op(input1: str): print("VertexAI pipeline: {}".format(input1))
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion
Now, you define the pipeline. The `experimental.run_as_aiplatform_custom_job` method takes as args the component defined above, and the list of `worker_pool_specs`— in this case one— with which the custom training job is configured. See [full function code here](https://github.com/kubeflow/pipelines/blob/master/sdk/p...
# Output directory and job_name OUTDIR = f"gs://{BUCKET}/taxifare/trained_model_{TIMESTAMP}" MODEL_DISPLAY_NAME = f"taxifare_{TIMESTAMP}" PYTHON_PACKAGE_URIS = f"gs://{BUCKET}/taxifare/taxifare_trainer-0.1.tar.gz" MACHINE_TYPE = "n1-standard-16" REPLICA_COUNT = 1 PYTHON_PACKAGE_EXECUTOR_IMAGE_URI = ( "us-docker.pk...
_____no_output_____
Apache-2.0
notebooks/building_production_ml_systems/labs/3_kubeflow_pipelines_vertex.ipynb
paras301/asl-ml-immersion