# Import required packages

# Import pandas in order to access NCEI data through its URL
import pandas as pd

# Import holoviews extension to pandas to make an interactive plot later
import holoviews as hv
import hvplot.pandas

# Create variable name for URL
## Create URL by generating API endpoint using NCEI [API documentation]
## (https://www.ncei.noaa.gov/support/access-data-service-api-user-documentation)
### Call variable by name at end of cell

st_petersburg_url = (
    'https://www.ncei.noaa.gov/access/services/data/v1'
    '?dataset=daily-summaries'
    '&dataTypes=TAVG'
    '&stations=RSM00026063'
    '&units=standard'
    '&startDate=1881-01-01'
    '&endDate=2023-12-31')
st_petersburg_url

'https://www.ncei.noaa.gov/access/services/data/v1?dataset=daily-summaries&dataTypes=TAVG&stations=RSM00026063&units=standard&startDate=1881-01-01&endDate=2023-12-31'

# Create expressive variable name for a dataframe (a type of Python object)
## Import data into Python using module 'pd' and function 'read.csv()'
## from NCEI API (AKA: download your data including parameters)
### Call dataframe variable at end of cell
sp_climate_df = pd.read_csv(
    st_petersburg_url,
    index_col="DATE",
    parse_dates=True,
    na_values=['NaN'])
sp_climate_df

# Check that the data was imported into a pandas DataFrame (object)
sp_climate_df.info()

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 52226 entries, 1881-01-01 to 2023-12-31
Data columns (total 2 columns):
 #   Column   Non-Null Count  Dtype 
---  ------   --------------  ----- 
 0   STATION  52226 non-null  object
 1   TAVG     52226 non-null  int64 
dtypes: int64(1), object(1)
memory usage: 1.2+ MB

# Clean up data - get rid of unwanted STATION column by selecting only 
# the columns I want which is [[TAVG]]
## Call DataFrame by name at end of cell
sp_climate_df = sp_climate_df[['TAVG']]
sp_climate_df

# Store the DataFrame created to use in other Notebooks
%store sp_climate_df

Stored 'sp_climate_df' (DataFrame)

# Label unit type by using the method .rename and the function (columns=) 
sp_climate_df_units = sp_climate_df.rename(columns={
    'TAVG': 'TAVG_F'
})

sp_climate_df_units

# Try plotting DataFrame to make sure what you are seeing makes sense
sp_climate_df.plot(
   y='TAVG',
   title='St.Petersbug, Russia - Average Temperature',
   xlabel='Year',
   ylabel='Temperature ($^\circ$F)'
)

<Axes: title={'center': 'St.Petersbug, Russia - Average Temperature'}, xlabel='Year', ylabel='Temperature ($^\\circ$F)'>

# Create an expressive new variable name for annual temp
## Resample (a type of method) data, setting the frequency of the final 
## data to year start ('YS') using .mean (another type of method)
### Call variable by name at end of cell
ann_sp_climate_df = sp_climate_df.resample('YS').mean()
ann_sp_climate_df

 # Plot this new variable
ann_sp_climate_df.plot(
    y='TAVG',
    title='St.Petersbug, Russia - Annual Average Temperature',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
)

<Axes: title={'center': 'St.Petersbug, Russia - Annual Average Temperature'}, xlabel='Year', ylabel='Temperature ($^\\circ$F)'>

# Try creating an interavtive plot 
ann_sp_temp_plot = ann_sp_climate_df.hvplot(
    y='TAVG',
    title='Annual Temperature in St.Petersburg, Russia Over Time',
    xlabel='Year',
    ylabel='Temperature Degrees Fahrenheit'
    )
ann_sp_temp_plot

/opt/conda/lib/python3.10/site-packages/holoviews/core/data/pandas.py:39: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
  return dataset.data.dtypes[idx].type
/opt/conda/lib/python3.10/site-packages/holoviews/core/data/pandas.py:39: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
  return dataset.data.dtypes[idx].type

# Save my work
hv.save(ann_sp_temp_plot, 'st_petersburg_ann_temp_plot.html')

/opt/conda/lib/python3.10/site-packages/holoviews/core/data/pandas.py:39: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
  return dataset.data.dtypes[idx].type
/opt/conda/lib/python3.10/site-packages/holoviews/core/data/pandas.py:39: FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`
  return dataset.data.dtypes[idx].type

%store ann_sp_climate_df

Stored 'ann_sp_climate_df' (DataFrame)

# Import packages needed to calculate and plot an OLS Linear trend line.
## Advanced options on matplotlib/seaborn/pandas plots
import matplotlib.pyplot as plt
# Common statistical plots for tabular data
import seaborn as sns
# Fit an OLS linear regression
from sklearn.linear_model import LinearRegression

# Reshape data to be 2D arrays for scikit-learn
predictor = ann_sp_climate_df.index.year.values.reshape(-1, 1)
observed = ann_sp_climate_df['TAVG'].values

# Create and fit the linear regression model
model = LinearRegression()
model.fit(predictor, observed)

# Get the slope and intercept
slope = model.coef_[0]
intercept = model.intercept_

# Print the results
print(f"Slope: {slope}")
print(f"Intercept: {intercept}")

Slope: 0.0304989502413193
Intercept: -18.536096426369156

fig = plt.figure(figsize=(10, 6))

# Plot annual average temperature data with a trend line
ax = sns.regplot(
    x=ann_sp_climate_df.index.year, 
    y=ann_sp_climate_df.TAVG,
    )
# Set plot labels
ax.set(
    title='St. Petersburg, Russia Annual Average Temperature, 1881-2023',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
)
# Display the plot without extra text
ann_sp_temp_trend=(
    plt.show()
)
fig.savefig('st_petersburg_ann_temp_trend_1881_2023.png')

fig = plt.figure(figsize=(10, 6))

# Clean the data to only include 1944 and beyond
ann_sp_climate_df_clean_1944 = (
    ann_sp_climate_df
    .loc['1944':]
)
# Plot this new clean data post 1943
ann_sp_climate_df_clean_1944.plot(
    y='TAVG',
    title='St.Petersbug, Russia - Annual Average Temperature, 1944 - 2023',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
    )
fig.savefig('st_petersburg_ann_temp_plot_1944_2023.png')

<Figure size 1000x600 with 0 Axes>

# Reshape data to be 2D arrays for scikit-learn
predictor = ann_sp_climate_df_clean_1944.index.year.values.reshape(-1, 1)
observed = ann_sp_climate_df_clean_1944['TAVG'].values

# Create and fit the linear regression model
model = LinearRegression()
model.fit(predictor, observed)

# Get the slope and intercept
slope = model.coef_[0]
intercept = model.intercept_

# Print the results
print(f"Slope: {slope}")
print(f"Intercept: {intercept}")

Slope: 0.05668220433083306
Intercept: -70.63108173262603

fig = plt.figure(figsize=(10, 6))

# Plot annual average temperature data with a trend line
ax = sns.regplot(
    x=ann_sp_climate_df_clean_1944.index.year, 
    y=ann_sp_climate_df_clean_1944.TAVG,
    )
# Set plot labels
ax.set(
    title='St. Petersburg, Russia Annual Average Temperature, 1944-2023',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
)
# Display the plot without extra text
plt.show()
fig.savefig('st_petersburg_ann_temp_trend_1944_2023.png')

fig = plt.figure(figsize=(10, 6))

# Clean data to be only 1990 and beyond
ann_sp_climate_df_clean_1990 = (
    ann_sp_climate_df
    .loc['1990':]
)
# Plot this new data for post 1989
ann_sp_climate_df_clean_1990.plot(
    y='TAVG',
    title='St.Petersbug, Russia - Annual Average Temperature, 1990 - 2023',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
    )
fig.savefig('st_petersburg_ann_temp_plot_1990_2023.png')

<Figure size 1000x600 with 0 Axes>

# Reshape data to be 2D arrays for scikit-learn
predictor = ann_sp_climate_df_clean_1990.index.year.values.reshape(-1, 1)
observed = ann_sp_climate_df_clean_1990['TAVG'].values

# Create and fit the linear regression model
model = LinearRegression()
model.fit(predictor, observed)

# Get the slope and intercept
slope = model.coef_[0]
intercept = model.intercept_

# Print the results
print(f"Slope: {slope}")
print(f"Intercept: {intercept}")

Slope: 0.05769689515732562
Intercept: -72.52459556040657

fig = plt.figure(figsize=(10, 6))

# Plot annual average temperature data with a trend line
ax = sns.regplot(
    x=ann_sp_climate_df_clean_1990.index.year, 
    y=ann_sp_climate_df_clean_1990.TAVG,
    )
# Set plot labels
ax.set(
    title='St. Petersburg, Russia Annual Average Temperature, 1990-2023',
    xlabel='Year',
    ylabel='Temperature ($^\circ$F)'
)
# Display the plot without extra text
plt.show()
fig.savefig('st_petersburg_ann_temp_trend_1990_2023.png')

	TAVG
DATE
1881-01-01	32
1881-01-02	30
1881-01-03	34
1881-01-04	35
1881-01-05	21
...	...
2023-12-27	24
2023-12-28	26
2023-12-29	27
2023-12-30	24
2023-12-31	13

	TAVG_F
DATE
1881-01-01	32
1881-01-02	30
1881-01-03	34
1881-01-04	35
1881-01-05	21
...	...
2023-12-27	24
2023-12-28	26
2023-12-29	27
2023-12-30	24
2023-12-31	13

	TAVG
DATE
1881-01-01	37.460274
1882-01-01	41.443836
1883-01-01	39.750685
1884-01-01	38.715847
1885-01-01	39.813699
...	...
2019-01-01	44.276712
2020-01-01	45.926230
2021-01-01	43.055249
2022-01-01	43.610959
2023-01-01	44.156164

Climate Portfolio Post - St.Petersburg, Russia¶

Site Description¶

Data Description¶

Citations for Data Description¶

Methods Overview¶

Workflow:¶

Interpretation Average Temperature Plot¶

Average Data in St. Petersburg while appears cooler, the overall range is appropriate given degrees fahrenheit¶

Interpretation of resampled, year start data¶

Annual Average Data in St. Petersburg increased between 1881 and 2023 and appears more drastic than the exercise with the Boulder data which is in line with my source for the site description¶

Interpretation of Interactive Average Annual Temperature Plot¶

Annual Average Data in St. Petersburg increased between 1881 and 2023 with steep shifts in 1941-1943 and 1987-1989 that need to be accounted for¶

Interpret the trend¶

Increase in Annual Average Temperature 1881-2023 in St. Petersburg Russia¶

while global annual temperatures also rise due to Gloabl Warming¶

Try looking at slope and trend line of data after 1943¶

Interpret the after 1943 plot¶

Annual Average Data in St. Petersburg increased between 1944 and 2023 and appears more drastic than the exercise with the Boulder data which is in line with my source for the site description¶

Interpret the Trend after 1943¶

Increase in Annual Average Temperature 1944-2023 in St. Petersburg Russia¶

is a Higher Increase in Temp Than the 1881-2023 Trend - while global annual¶

temperatures also rose due to Gloabl Warming¶

Try looking at slope and trend line of data after 1989¶

Interpret the plot¶

Annual Average Data in St. Petersburg increased between 1881 and 2023 and appears more drastic than the exercise with the Boulder data which is in line with my source for the site description¶

Interpret the Trend after 1989¶

Increase in Annual Average Temperature 1990-2023 in St. Petersburg Russia is a Higher Increase in Temp Than the 1881-2023 Trend and only slightly higher than 1944-2023 - while global annual temperatures also rose due to Gloabl Warming¶

Conclusion¶

	STATION	TAVG
DATE
1881-01-01	RSM00026063	32
1881-01-02	RSM00026063	30
1881-01-03	RSM00026063	34
1881-01-04	RSM00026063	35
1881-01-05	RSM00026063	21
...	...	...
2023-12-27	RSM00026063	24
2023-12-28	RSM00026063	26
2023-12-29	RSM00026063	27
2023-12-30	RSM00026063	24
2023-12-31	RSM00026063	13