Application Programming Interfaces

Task 1

wake a response to the Mensa-Api for the timeframe today + 1 Days and today + 7 Days and extract the current Data for meals. Save your data in the Variable meals.

Get today:

today = date.today()
today

datetime.date(2023, 12, 22)

Define URL:

API_URL = "https://sls.api.stw-on.de/v1/location/101/menu/{}/{}".format(
    today + timedelta(days = 1), today + timedelta(days = 7))
API_URL

'https://sls.api.stw-on.de/v1/location/101/menu/2023-12-23/2023-12-29'

Wake a response and convert to json:

resp = requests.get(API_URL)
resp_json = resp.json()

Extract Meals Data:

meals = resp_json.get("meals")
meals[0]['price']

{'student': '2.05', 'employee': '3.75', 'guest': '4.85'}

Now convert the list of meals into a Pandas DataFrame object and store it in df_meals. Make sure that the date column has the datetime data type and that the prices are in a numeric (float) format.

Hint: You might want to check the pandas.json_normalize function.

df_meals = pd.json_normalize(meals)
df_meals["date"] = pd.to_datetime(df_meals["date"])
df_meals["price.student"] = pd.to_numeric(df_meals["price.student"])
df_meals.head(2)

	id	date	name	name_en	time	special_tags	price.student	price.employee	price.guest	nutritional_values._NOTE	...	location.address.zip	location.address.city	location.opening_hours	lane.id	lane.name	lane.name_en	tags.categories	tags.allergens	tags.additives	tags.special
0	50181	2023-12-18	Paprikasuppe	Bell pepper cream soup	noon	[Deprecated. Use tags→special instead.]	2.05	3.75	4.85	WARNING: These fields currently contain incorr...	...	38106	Braunschweig	[{'time': 'noon', 'start_day': 1, 'end_day': 4...	10	Suppe & Co.	Soup & Co.	[{'id': 'VEGA', 'name': 'Vegan', 'name_en': 'v...	[{'id': 'SO', 'name': 'enthält Soja(bohnen)u d...	[{'id': '2', 'name': 'mit Konservierungsstoff'...	[]
1	50183	2023-12-18	Bulgurpfanne mit Wildkräutern | Kräuterjoghurt...	Bulgur with wild herbs and courgettes | Herb y...	noon	[Deprecated. Use tags→special instead.]	2.50	5.80	6.90	WARNING: These fields currently contain incorr...	...	38106	Braunschweig	[{'time': 'noon', 'start_day': 1, 'end_day': 4...	20	Classic 1	Classic 1	[{'id': 'VEGT', 'name': 'Vegetarisch', 'name_e...	[{'id': 'ML', 'name': 'enthält Milch u Milcher...	[]	[]

2 rows × 33 columns

Next, create a simple plot that plots the student prices over time. You can use a scatter plot to do this.

plt.scatter(df_meals["date"], df_meals["price.student"])
plt.show()

Task 2

Next, make a request to the Mensa API to get all meals in the time frame December 1, 2023 to December 21, 2023.

After requesting the data, Analyze it by their mean price Distribution, and a thing you find interesting to analyze. (Be Creative)

What can you observe?

API_URL = "https://sls.api.stw-on.de/v1/location/101/menu/2023-12-01/2023-12-21"

resp = requests.get(API_URL)
data = resp.json()

meals=data.get("meals")
len(meals)

40

df_meals=pd.json_normalize(meals, max_level=1)
df_meals=df_meals.astype({'price.student': 'float',"price.employee":"float", "price.guest":"float"})
df_meals["date"]=df_meals["date"].astype('datetime64[ns]')
df_meals.head(1)

	id	date	name	name_en	time	special_tags	price.student	price.employee	price.guest	nutritional_values._NOTE	...	location.name	location.address	location.opening_hours	lane.id	lane.name	lane.name_en	tags.categories	tags.allergens	tags.additives	tags.special
0	50181	2023-12-18	Paprikasuppe	Bell pepper cream soup	noon	[Deprecated. Use tags→special instead.]	2.05	3.75	4.85	WARNING: These fields currently contain incorr...	...	Mensa 1 TU Braunschweig	{'line1': 'Mensa 1 TU Braunschweig', 'line2': ...	[{'time': 'noon', 'start_day': 1, 'end_day': 4...	10	Suppe & Co.	Soup & Co.	[{'id': 'VEGA', 'name': 'Vegan', 'name_en': 'v...	[{'id': 'SO', 'name': 'enthält Soja(bohnen)u d...	[{'id': '2', 'name': 'mit Konservierungsstoff'...	[]

1 rows × 22 columns

Analyzing Price Distribution:

aggregation_functions = {'date': 'first',
                         'id': 'first',
                         'price.student': 'mean',
                         'price.employee':"mean",
                         "price.guest":"mean",
                         "location.id":"first",
                         "lane.id":"first"}
df_new = df_meals.groupby(df_meals['date']).aggregate(aggregation_functions)
df_new

	date	id	price.student	price.employee	price.guest	location.id	lane.id
date
2023-12-18	2023-12-18	50181	2.020	3.185	3.700	101	10
2023-12-19	2023-12-19	50189	1.975	3.235	3.715	101	10
2023-12-20	2023-12-20	50397	2.020	3.245	3.760	101	10
2023-12-21	2023-12-21	50225	2.135	3.300	3.815	101	10

#mean_price.plot
plt.plot(df_new["date"], df_new["price.student"], color="yellow",label="student")  
plt.plot(df_new["date"], df_new["price.employee"],color= "red", label="employee")
plt.plot(df_new["date"], df_new["price.guest"], color= "purple", label="guest")
plt.title('Prices Mensa over time')  
plt.xlabel('Date')  
plt.ylabel('Price')  
plt.show()

Data Analysis

Setup

import pandas as pd
import matplotlib.pyplot as plt

Next read in the dataset survey.csv. For testing puposes use the Variable Name df.

df = pd.read_csv("survey.csv")
df.head()

	Age	Sex	Scale Python Exp	Course	Has Voice Assistent Contact	Voice Assistent	Scale Study Satisfaction	Uses Smartphone	Which Smartphone	Has Computer	Which OS	Scale Programming Exp
0	22	Männlich	4	Medienwissenschaften	Ja	Apple Siri	4	Ja	Apple	Ja	Mac OS	2
1	26	Weiblich	3	Medienwissenschaften	Ja	Amazon Alexa	2	Ja	Xiaomi	Ja	Windows 10	3
2	21	Männlich	3	Medienwissenschaften	Ja	Google Now	4	Ja	Sonstige	Ja	Windows 10	3
3	26	Weiblich	4	Medienwissenschaften	Ja	Apple Siri	4	Ja	Samsung	Ja	Windows 10	2
4	24	Weiblich	4	Psychologie	Nein	NaN	4	Ja	Apple	Ja	Windows 11	3

Hypothesis 1

The average age in the course is 25.34 years, with a surplus of females.

print("Mean Age:", df["Age"].mean())
c = df["Sex"].value_counts()
plt.title("Gender Distribution")
plt.bar(c.keys(), c, color=["black", "white"])
plt.show()

Mean Age: 22.64

Hypothesis 2

The most used voice assistant is Alexa.

c = df["Voice Assistent"].value_counts()
plt.title("Voice Assistent Distribution")
plt.bar(c.keys(), c, color=["red", "green", "blue"])
plt.show()

Hypothesis 3

Hypothesis 3.1

The least used smartphone operating system is iOS.

os_dict = {
    "iOS": 0,
    "Android": 0
}

for data in df["Which Smartphone"]:
    if data == "Apple":
        os_dict["iOS"] += 1
    else:
        os_dict["Android"] += 1

plt.title("Mobile OS Distribution")
plt.bar(os_dict.keys(), os_dict.values(), color=["silver", "green"])
plt.show()

Hypothesis 3.2

The most used desktop operating system is Windows 10.

os_dict = df["Which OS"].value_counts()
plt.title("Desktop OS Distribution")
plt.bar(os_dict.keys(), os_dict, color=["grey", "silver", "cyan", "yellow"])
plt.show()

Hypothesis 4

Hypothesis 4.1 & 4.2

The youngest people use an iPhone. Older people use an Android-based smartphone.

age_mean = df["Age"].mean()
youngest = {
    "ios": 0,
    "Android": 0,
    "n": 0,
    "ratio iOS": 0,
    "ratio Android": 0
}

oldest = {
    "ios": 0,
    "Android": 0,
    "n": 0,
    "ratio iOS": 0,
    "ratio Android": 0
}

Count appreances:

for index, row in df[["Age", "Which Smartphone"]].iterrows():
    if age_mean > row["Age"]:
        youngest["n"] += 1
        if row["Which Smartphone"] == "Apple":
            youngest["ios"] += 1
        else:
            youngest["Android"] += 1

    if age_mean < row["Age"]:
        oldest["n"] += 1
        if row["Which Smartphone"] != "Apple":
            oldest["Android"] += 1
        else:
            oldest["ios"] += 1

Calc Ratios:

youngest["ratio iOS"] = youngest["ios"] / youngest["n"] * 100
youngest["ratio Android"] = youngest["Android"] / youngest["n"] * 100

oldest["ratio iOS"] = oldest["ios"] / oldest["n"] * 100
oldest["ratio Android"] = oldest["Android"] / oldest["n"] * 100

plt.title("iOS Ratio between mean Age")
plt.bar(["Y: mean iOS", "Y: mean Android", "O: mean iOS", "O: mean Android"], 
        [youngest["ratio iOS"], youngest["ratio Android"], oldest["ratio iOS"], oldest["ratio Android"]], color=["gold", "silver"])
plt.show()