1. CIT 590 Intro to Programming Lecture 6 (dictionaries)

2. Agenda Dictionary

3. Dictionaries consist of key, value pairs Also know as Hashmaps, associative arrays in other languages Initialized with d = {} Do not use the variable name dict. ‘reserved’ words Very efficient way of storing sparse data A lot of matrices and vectors that come up in probability are sparse, so you could use an integer key and store values in that manner

4. Simple operations The standard len and del work Loops work over the keys of the dictionary Dictionary.get()

5. Dictionary update operation. Copying dictionaries Dictone = [‘abc’: 3, ‘def’:7, ‘xyz’: 9} Dicttwo = [‘def’: 5, ‘pdq’ : 4} Dictone.update(dicttwo) As with lists or for that matter any kind of assignment, you need to be careful when you do assignment, since that is done by reference So dict1 = dict2 will make both dictionaries refer to the same value If you want to make what is generally called a ‘shallow copy’, you need to use the copy method Dict 2 = dict1.copy()

6. Initializing a dictionary version duex The dict() function can convert a list of 2 element tuples into dictionaries Very useful when you want to initialize a dictionary by using 2 lists, one of which has keys = [‘Arvind’, ‘Aashish’] values = [33, 28] dict(zip(keys, values))

7. Looping over a dictionary For e in dict: This will loop over they keys For e in dict.values(): As is somewhat obvious, this will loop over the values Since there is no guarantee about what order the keys are looped over, you might sometimes want to call the sorted function

8. Shallow copy versus deep copy Difference comes up in compound objects A shallow copy constructs a new compound object and then (to the extent possible) inserts references into it to the objects found in the original. A deep copy constructs a new compound object and then, recursively, inserts copies into it of the objects found in the original You can manage to do both on a dictionary if you import the copy module Copy.copy = shallow copy Copy.deepcopy = deep copy Example with matrices

9. Simple program samples duesProgram.py The second optional argument for get can be exploited in interesting ways Useful for maintaining records Concordance.py

10. Altering dictionaries by passing them into functions Remember that arguments are passed into functions by reference, so you could get some unintended consequences if you are not careful Similar to our sneakyAppend examples in lists The debugger can be extremely useful here since it will always show you the complete dictionary

11. Dynamic programming The central principle is that you delay the computation of the function as much as possible When you compute a result, store it somewhere in memory Retrieve the results from memory Watch recursive fibonacci suddenly not become such a beast. Yes, this does assume we have memory (for a lot of cases this is ok)

12. Persistent variables The shelve module allows data to persist without using explicit file commands Data persists in a dictionary! import shelve Data = shelve.open(“database”) #and this will make an implicit file called database And then data from that point can basically be treated just as any other dictionary Remember to close the shelf before quitting by using Data.close() Files without really using files

13. The internal dictionaries Locals() gives you the local variables Correspondingly global() gives you the global variables

14. Summary Read chapter 5 in its entireity