Lecture 6 / Lab 2 - Answering Weather Questions

• Announcements:

  • Quiz 1 results out after class
    • Some data type questions had multiple valid answers
  • Quiz 2 this weekend - take it before Monday's class!
  • Lab 1 Survey didn't have the open-ended feedback question. I added it and will accept any late survey submissions if you're willing to go back and give me feedback!
  • Lab 2 out today, due next Thursday 10pm.

• Goals:

  • Know how to make basic scatter plots to look at bivariate relationships of variables in a Pandas data frame
  • Know how to perform basic filtering by time and date.
  • Know how to create compound boolean masks using & and | (beware the precedence!)
  • Get started on Lab 2

def data_url(city_name):
    base_url = "https://facultyweb.cs.wwu.edu/~wehrwes/courses/csci141_21s/fp/data/"
    return base_url + city_name + ".csv"

Game Plan:

• Load bellingham's weather data.
• Clean and convert HourlyDryBulbTemperature and HourlyPrecipitation to floats
• Convert DATE from object/string to a datetime
• Extract only the observations from February
• Make histograms of the above variables individually
• Make a scatter plot that shows the relationship between the above variables.

import pandas as pd

def data_url(city_name):
    base_url = "https://facultyweb.cs.wwu.edu/~wehrwes/courses/csci141_21s/fp/data/"
    return base_url + city_name + ".csv"

bham = pd.read_csv(data_url("WA_Bellingham"), low_memory=False)

# Convert HourlyDryBulbTemperature to float, stripping 's' off the end where applicable
temps = bham["HourlyDryBulbTemperature"].str.rstrip("s").astype(float)
bham["HourlyDryBulbTemperature"] = temps

# Convert HourlyPrecipitation to float, replacing "T" with 0 and stripping "s"
precips = bham["HourlyPrecipitation"].str.replace("T", "0").str.rstrip("s")
bham["HourlyPrecipitation"] = precips.astype(float)
bham[["DATE", "HourlyDryBulbTemperature", "HourlyPrecipitation"]].info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 11348 entries, 0 to 11347
Data columns (total 3 columns):
DATE                        11348 non-null object
HourlyDryBulbTemperature    10982 non-null float64
HourlyPrecipitation         9296 non-null float64
dtypes: float64(2), object(1)
memory usage: 266.0+ KB

• Convert DATE from object/string to a datetime

bham["DATE"] = pd.to_datetime(bham["DATE"])
bham[["DATE", "HourlyDryBulbTemperature", "HourlyPrecipitation"]]

	DATE	HourlyDryBulbTemperature	HourlyPrecipitation
0	2020-01-01 00:14:00	53.0	NaN
1	2020-01-01 00:24:00	53.0	NaN
2	2020-01-01 00:53:00	53.0	0.00
3	2020-01-01 01:53:00	52.0	0.00
4	2020-01-01 02:53:00	52.0	0.00
5	2020-01-01 03:53:00	52.0	0.00
6	2020-01-01 04:53:00	51.0	NaN
7	2020-01-01 05:53:00	50.0	NaN
8	2020-01-01 06:53:00	49.0	0.00
9	2020-01-01 07:53:00	49.0	0.00
10	2020-01-01 08:53:00	49.0	0.00
11	2020-01-01 09:53:00	50.0	0.00
12	2020-01-01 10:53:00	50.0	0.00
13	2020-01-01 11:53:00	51.0	0.00
14	2020-01-01 12:53:00	50.0	0.00
15	2020-01-01 13:53:00	52.0	NaN
16	2020-01-01 14:53:00	50.0	0.00
17	2020-01-01 15:53:00	50.0	0.00
18	2020-01-01 16:53:00	50.0	0.00
19	2020-01-01 17:53:00	49.0	0.00
20	2020-01-01 18:53:00	48.0	0.00
21	2020-01-01 19:53:00	46.0	0.00
22	2020-01-01 20:53:00	45.0	0.00
23	2020-01-01 21:27:00	45.0	NaN
24	2020-01-01 21:53:00	45.0	0.00
25	2020-01-01 22:53:00	44.0	0.00
26	2020-01-01 23:53:00	44.0	0.00
27	2020-01-01 23:59:00	NaN	NaN
28	2020-01-02 00:53:00	44.0	0.00
29	2020-01-02 01:53:00	44.0	0.00
...	...	...	...
11318	2020-12-31 06:53:00	41.0	0.02
11319	2020-12-31 07:10:00	41.0	0.00
11320	2020-12-31 07:53:00	41.0	0.00
11321	2020-12-31 08:53:00	41.0	0.00
11322	2020-12-31 09:18:00	42.0	NaN
11323	2020-12-31 09:39:00	42.0	NaN
11324	2020-12-31 09:53:00	42.0	0.00
11325	2020-12-31 10:53:00	43.0	0.00
11326	2020-12-31 11:53:00	43.0	0.00
11327	2020-12-31 12:53:00	44.0	0.00
11328	2020-12-31 13:53:00	45.0	0.00
11329	2020-12-31 14:53:00	45.0	NaN
11330	2020-12-31 15:53:00	45.0	0.01
11331	2020-12-31 16:53:00	44.0	0.01
11332	2020-12-31 16:56:00	44.0	NaN
11333	2020-12-31 17:15:00	45.0	0.00
11334	2020-12-31 17:38:00	45.0	0.00
11335	2020-12-31 17:53:00	45.0	0.01
11336	2020-12-31 17:55:00	45.0	0.00
11337	2020-12-31 18:02:00	45.0	0.00
11338	2020-12-31 18:09:00	45.0	0.00
11339	2020-12-31 18:16:00	45.0	0.00
11340	2020-12-31 18:43:00	45.0	0.00
11341	2020-12-31 18:53:00	45.0	0.00
11342	2020-12-31 19:53:00	46.0	NaN
11343	2020-12-31 20:53:00	49.0	NaN
11344	2020-12-31 21:53:00	50.0	NaN
11345	2020-12-31 22:53:00	49.0	0.01
11346	2020-12-31 23:53:00	49.0	0.03
11347	2020-12-31 23:59:00	NaN	NaN

11348 rows × 3 columns

• Extract only the observations from February

mask = (bham['DATE'] > "2020-02-01") & (bham['DATE'] < "2020-03-01")
feb = bham[mask]

• Make histograms of the above variables individually

feb["HourlyDryBulbTemperature"].plot.hist()

<matplotlib.axes._subplots.AxesSubplot at 0x7fe756c28588>

feb["HourlyPrecipitation"].plot.hist()

<matplotlib.axes._subplots.AxesSubplot at 0x7fe76f94cc18>

• Make a scatter plot that shows the relationship between the above variables.

feb.plot.scatter("HourlyDryBulbTemperature", "HourlyPrecipitation")

<matplotlib.axes._subplots.AxesSubplot at 0x7fe76f8dd860>