Class 3: Booleans, Dictionaries, and Mammalian Diet#

EEB 125, W2025#

Tomo Parins-Fukuchi#

Concepts and tools#

  • Dictionaries

  • List comprehensions

  • Cleaning data and finding patterns

Some more techniques/tools in Python#

Dictionaries#

  • Another type of useful Python container is called a ‘dictionary’

  • They are specified using curly brackets: {}

  • Each dictionary contains a ‘key’ and a ‘value’. Let’s explore:

# create an empty dictionary and assign it to a variable
test_dict = {}
print(test_dict)

# create an entry in our new dictionary:
test_dict["test_key"] = "test_value"
print(test_dict)

# we can use the key to access the value:
print(test_dict["test_key"])

Dictionaries#

  • Keys and values can be almost any data type

  • Keys have some restrictions on data type, but strings, floats, etc, all are fine (more later)

# ex. 1. what data type are these keys? the values?
test_dic = {"a" : [1.3,2.2], "b" : [4.3]}
print(test_dic["a"])

# ex. 2. what data type are these keys? the values?
test_dic = {0 : ["a","b"], 1 : ["c"]}
print(test_dic[0])

# ex. 3. what data type are these keys? the values?
test_dic = {0 : ["a","b"], "z" : ["c"]}
print(test_dic["z"])

# dictionaries offer a lot of flexibility

test_dic[0].append("d")
print(test_dic[0])

# we can also loop over a dictionary

testd = {"21" : "ia>iw", "kendrick" : "gkmc","weezy":"tc3"}
for mc in testd:
    key = mc
    value = testd[mc]
    print(key,value)

List comprehensions#

  • Python offers another way of iterating through data that can be handy

  • It is similar to a for loop, but stated more compactly

# count from zero to nine using a for loop and store the results in a list

zero_to_nine = []
for i in range(10):
    zero_to_nine.append(i)
print(zero_to_nine)

# the same thing using a list comprehension

zero_to_nine = [i for i in range(10)]
print(zero_to_nine)

List comprehensions#

  • We can use these to create slightly modified versions of existing lists

emcees_str = "cole, rocky, 21, weezy, kendrick"
emcees_ls = emcees_str.split(",")

# we can use a list comprehension to generate a cleaner version of emcees_ls

emcees_ls_cln = [mc.strip().upper() for mc in emcees_ls]
print(emcees_ls_cln)

A brief aside#

A brief aside#

Which Canadian province has the most fossils?#

We can use list comprehensions to clean up the results we arrived at last week.

file = open("pbdb_data.csv")
file

lines = file.readlines()
header = lines[0]
data = lines[1:]

## review from last week

province_ls = ["Ontario","British Columbia","Alberta","Saskatchewan","Manitoba","Newfoundland and Labrador","Northwest Territories","Yukon","Prince Edward Island","Nunavut"]

province_recs = []
for line in data:
    line_dat = line.strip().split(",")
    province = line_dat[-1].strip()
    province_recs.append(province)

# RESULTS FROM LAST WEEK

for i in province_ls:
    print(i,province_recs.count(i))

Finding unique values#

  • We might want to pick out the unique values in a data column

    • We don’t always know what categories our data is divided into

  • Python set does this naturally

Python set data structure#

  • A set is a native data container (a listis another)

    • No duplicates

    • We can use it to find the unique values within our data

genres = ["black metal","rap","rap","country","black metal","rap"]
print(genres)
print(set(genres))

# Notice a few of these seem to just be different in caps (e.g., "British columbia")

unique = set(province_recs)
print(unique)

province_ls_uc = [prov.upper() for prov in province_ls]
province_recs_uc = [prov.upper() for prov in province_recs]
for i in province_ls_uc:
    print(i,province_recs_uc.count(i))

How many taxa are found in the fossil records of each province?#

We can use these new tools to ask another data science question: which Canadian province (or territory) has the most diverse fossil record?

Species richness#

  • Many questions in ecology and evolution start by asking how many species exist in a particular area

    • Fossils are often difficult to identify to the species level

    • We can use ‘taxon’ richness– could be species or could be higher taxonomic categories

Linnaean taxonomy#

Our ‘taxon richness’ metric could include any of these levels

print(header)

Lists as dictionary values#

  • We can create a dictionary to store all of the taxa found in each province

  • Start by creating an empty dictionary and fill it up

    • unique province records as keys

    • empty list as values

prov_tax = {}
for prov in unique:
    upperprov = prov.upper()
    prov_tax[upperprov] = []

Populate dictionary#

  • Loop over the lines of our data file

  • For each line, add the taxon name to the list associated with the province

for line in lines[1:]:
    line_dat = line.strip().split(",")
    prov = line_dat[-1].upper().strip()
    tax = line_dat[3]
    prov_tax[prov].append(tax)

Calculate taxon richness for each province#

  • Look up the provinces in our dictionary, prov_tax

  • Count the number of taxa associated with each one

print("province,num_taxa")
for prov in province_ls_uc: # why iterate over province_ls_uc instead of prov_tax?
    tax_ls = prov_tax[prov]
    unique_tax = set(tax_ls)
    num_tax = len(unique_tax)
    print(f"{prov},{num_tax}")

Switching gears#

  • Now, we will introduce a different data set and explore some more questions using some new concepts

  • But first, a break

Trophic levels#

Alt Text

  • Sort of an oversimplification

  • But generally answers “does it eat plants” vs “does it eat herbivores” vs “does it eat carnivores”

  • There are no primary producers among mammals (or any other animal)

Trophic levels in Mammals#

  • We will now explore a new data science question:

    • What is the distribution of trophic levels in different mammalian taxa?

    • Will explain this more as we go

file = open("trophic_level.csv")
lines = file.readlines()
header = lines[0]
data = lines[1:]

print(header)

print(lines[1])

# lets try to see what levels are in this dataset

levels = []
for line in data:
    dat = line.strip().split(",")
    level = dat[-1]
    levels.append(level)

What do you suppose these mean?

unique = set(levels)
print(unique)

Trophic level#

Trophic level of each species measured using any qualitative or quantitative dietary measure, over any period of time, using any assessment method, for non-captive or non-provisioned populations; adult or age unspecified individuals, male, female, or sex unspecified individuals; primary, secondary, or extrapolated sources; all measures of central tendency; in all localities. Species were defined as (1) herbivore (not vertebrate and/or invertebrate), (2) omnivore (vertebrate and/or invertebrate plus any of the other categories) and (3) carnivore (vertebrate and/or invertebrate only)

Do bats or Carnivorans have more meat-eaters?#

  • Let’s shift to ask a more targetted question

    • Do bats have fewer or more meat-eating species than Carnivorans (dogs, cats, etc)?

    • How might we begin to approach this question??

# Q: how can we check how many rows there are in the dataset?

Selecting Rows using a Condition#

  • Okay, so we have a lot of rows to deal with. Sorting through manually would be impossible

  • We can select rows we want using Boolean logic

George Boole#

  • Self-taught mathematician from the 19th century

  • Pioneered (what we now call) Boolean algebra

    • System of algebra/logic using variables of True and False rather than numbers

    • Absolutely fundamental concept for the computational sciences

Alt Text

Boolean variables#

  • Is a condition True or False?

  • Python has notation for checking this:

    • a == b -> “a is equal to b”

    • a != b -> “a is not equal to b”

# We can compare numeric values:
1 == 1

# We can compare numeric values:
1 != 1

# We can compare numeric values:
1 == 2

# We can compare numeric values:
1 != 2

# We can also compare strings:
"star wars" == "star trek"

# We can also compare strings:
"star wars" != "star trek"

Booleans#

  • Booleans are another native data type in Python

type(True)

Using Booleans to extract data#

We can find the rows that correspond to each of our orders using Boolean logic

  • Find rows that return True when the order variable is equal to (e.g.) Bats

  • Can use if keyword in Python:

have_code_problems = False

if have_code_problems == True:
    print("i feel bad for ya son")
elif have_code_problems == False:
    print("noice")

Using Booleans to extract data#

  • Can use if statements to compare many types of data

prof = "KP"

if prof == "KP":
    print("wow, the semester is flying by!!!")
elif prof == "michael":
    print("i'm so glad that midterm is over, phew")
else:
    print("we must still be stuck with tomo...")

Preparing#

  • First, let’s find out how many species belong to each of our orders of interest

  • Loop through our data and simply count how many records correspond to each

n_bats = 0
n_carn = 0
for line in data:
    line_dat = line.strip().split(",")
    order = line_dat[0]
    if order == "Chiroptera":
        n_bats = n_bats + 1
    elif order == "Carnivora":
        n_carn = n_carn + 1
        
print("chiroptera:",n_bats)
print("carnivora:",n_carn)

Which bats eat meat? Which Carnivorans do?#

  • We can hone in on our focal question by just adding one more Boolean:

bats_meat = 0
carn_meat = 0
for line in data:
    line_dat = line.strip().split(",")
    order = line_dat[0]
    troph = line_dat[2]
    if order == "Chiroptera":
        if troph == "3":
            bats_meat = bats_meat + 1
    elif order == "Carnivora":
        if troph == "3":
            carn_meat = carn_meat + 1
        
print("chiroptera:",bats_meat)
print("carnivora:",carn_meat)

Which bats eat meat? Which Carnivorans do?#

  • We can hone in on our focal question by just adding one more Boolean

  • We can make this a little easier on the eyes by combining Booleans:

bats_meat = 0
carn_meat = 0
for line in data:
    line_dat = line.strip().split(",")
    order = line_dat[0]
    troph = line_dat[2]
    if order == "Chiroptera" and troph == "3":
            bats_meat = bats_meat + 1
    elif order == "Carnivora" and troph == "3":
            carn_meat = carn_meat + 1
        
print("chiroptera:",bats_meat)
print("carnivora:",carn_meat)

What proportion of each eats meat?#

prop_bats_meat = bats_meat / n_bats
prop_carn_meat = carn_meat / n_carn
print("chiroptera:",prop_bats_meat)
print("carnivora:",prop_carn_meat)

BUT WAIT#

Missing data#

  • Often times, our observations may be incomplete

  • These lead to ‘missing data’

  • One common way of representing missing data is by writing ‘NA’

# what can we do with these 'NA' things!??!

unique = set(levels)
print(unique)

Missing data#

  • In our loop, if we encounter a line where tropic_level is “NA”, tell Python to skip to the next line

  • Can use the continue keyword for this:

for i in range(5):
    if i == 2:
        continue
    print(i)

Missing data#

  • In our loop, if we encounter a line where tropic_level is “NA”, tell Python to skip to the next line

  • Can use the continue keyword for this:

n_bats = 0
n_carn = 0
for line in data:
    line_dat = line.strip().split(",")
    order = line_dat[0]
    troph = line_dat[2]
    if troph == "NA": 
        continue        # if the trophic level in the dataset is missing, skip to the next line without counting
    if order == "Chiroptera":
        n_bats = n_bats + 1
    elif order == "Carnivora":
        n_carn = n_carn + 1
        
print("chiroptera:",n_bats)
print("carnivora:",n_carn)

prop_bats_meat = bats_meat / n_bats
prop_carn_meat = carn_meat / n_carn
print("chiroptera:",prop_bats_meat)
print("carnivora:",prop_carn_meat)

BUT WAIT#

Trophic level#

Trophic level of each species measured using any qualitative or quantitative dietary measure, over any period of time, using any assessment method, for non-captive or non-provisioned populations; adult or age unspecified individuals, male, female, or sex unspecified individuals; primary, secondary, or extrapolated sources; all measures of central tendency; in all localities. Species were defined as (1) herbivore (not vertebrate and/or invertebrate), (2) omnivore (vertebrate and/or invertebrate plus any of the other categories) and (3) carnivore (vertebrate and/or invertebrate only)

Trophic level#

  • Maybe we want to consider anything at either levels 2 or 3 as meat-eating


Expanding our Boolean toolkit#

  • We can use the greater than (>) operator to accomplish this

10 > 1

10 >= 10

10 < 1

Expanding our Boolean toolkit#

  • Identify any species where trophic level is >= 2 as a meat-eater

bats_meat = 0
carn_meat = 0
for line in data:
    line_dat = line.strip().split(",")
    order = line_dat[0]
    troph = line_dat[2]
    if troph == "NA": 
        continue        # if the trophic level in the dataset is missing, skip to the next line without counting
    troph_int = int(troph)
    if order == "Chiroptera" and troph_int >= 2:
            bats_meat = bats_meat + 1
    elif order == "Carnivora" and troph_int >= 2:
            carn_meat = carn_meat + 1
        
print("chiroptera:",bats_meat)
print("carnivora:",carn_meat)

prop_bats_meat = bats_meat / n_bats
prop_carn_meat = carn_meat / n_carn
print("chiroptera:",prop_bats_meat)
print("carnivora:",prop_carn_meat)

END OF WEEK 3#