def get_simpleplot(savepath, forecasts, obs_week, week):
"""Get Simple plot:
------------------------------------
This function plots a simple line plot of student's weekly averaged
forecast for a week
------------------------------------
Parameters:
forecasts = array
provides weekly forecasted flow of each student
obs_week = float
week's observed flow
week = string
Week number for the forecast (1 or 2)
------------------------------------
Outputs: figure of simple line plot
"""
# Get the array of firstnames for the plot
firstnames = ef.getFirstNames()
fig3, ax = plt.subplots()
fig3.set_size_inches(10, 4)
clean_forecasts = [x for x in forecasts if not np.isnan(x)]
class_avg = np.mean(clean_forecasts)
simple_plot = ax.plot(forecasts, '-g', label='Forecast', alpha=.8)
plt.axhline(y=class_avg,
linestyle='dashed',
label='Class Avg',
alpha=.8,
color='red')
plt.axhline(y=obs_week,
linestyle='dotted',
label='Observed',
alpha=.8,
color='blue')
plt.xticks(ticks=np.arange(0, 19, 1), labels=firstnames, rotation=60)
title_string = 'Week ' + str(week) + ' Forecasts'
ax.set(title=title_string,
xlabel="Students",
ylabel="Weekly Avg Flow [cfs]")
ax.legend(fancybox=True, framealpha=1, shadow=True, borderpad=1)
fig3.patch.set_facecolor('xkcd:white')
plt.tight_layout()
plt.show()
fig3.savefig(savepath)
return simple_plot
def last_2_weeks_diff(savepath, obs_week, oneweek_forecasts, twoweek_forecasts,
bar_width):
# Get the array of firstnames for the plot
firstnames = ef.getFirstNames()
class_forecasts = pd.DataFrame(
{
'oneweek_forecasts': oneweek_forecasts,
'twoweek_forecasts': twoweek_forecasts
},
index=firstnames,
columns=['oneweek_forecasts', 'twoweek_forecasts'])
class_forecasts.insert(
2, 'Diff_1', np.array(class_forecasts['oneweek_forecasts'] - obs_week),
True)
class_forecasts.insert(
3, 'Diff_2', np.array(class_forecasts['twoweek_forecasts'] - obs_week),
True)
# Plotting Diff
stu = np.arange(0, 19, 1)
fig14 = plt.figure()
fig14.set_size_inches(25, 8)
ax = fig14.add_subplot()
w = bar_width
plt.xticks(stu + w / 2, firstnames, rotation=60, fontsize=15)
ax.bar(stu, class_forecasts.Diff_1, width=w, align='center', label='week1')
ax.bar(stu + w,
class_forecasts.Diff_2,
width=w,
align='center',
label='week2')
ax.axhline(y=0, linewidth=2, linestyle='--', color='k')
plt.xlabel('Student', fontsize=15)
plt.ylabel('Average Flow', fontsize=15)
ax.legend(loc='lower center',
fontsize=20,
bbox_to_anchor=(.5, -0.4),
ncol=5)
# plt.text(0, 0, 'Observed Flow', fontsize=21)
fig14.patch.set_facecolor('xkcd:white')
plt.tight_layout()
plt.show()
fig14.savefig(savepath)
# %%
import pandas as pd
import sys
sys.path.append('../')
import eval_functions as ef
import random
# ENTER INFO HERE
weekly_winners = ["Diana", "Alexa", "Ben", "Jake"]
evaluators = ["Adam", "Jill"]
# %%
# get list of students in the class using functions
# names = ef.getLastNames()
firstnames = ef.getFirstNames()
nstudent = len(firstnames)
# %%
# remove ineligible persons
not_eligible = weekly_winners + evaluators
for i in range(len(not_eligible)):
firstnames.remove(not_eligible[i])
firstnames
# %%
# create list of emojis and shuffle randomly
# SHAPE (# of list items) MUST MATCH FIRSTNAMES LIST
emoji_list = [
"🥇", "🥈", "🥉", "😭", "😭",
def plot_seasonal_rmse(savepath, seasonal_rmse):
"""Seasonal Root Mean Square Error:
-----------------------------------
This function plots the root mean square error of seasonal flow
data for week 1 to the most recent entry.
You have the option of entering in minumum and maximum y values.
-----------------------------------
Parameters:
Seasonal_rmse = pandas dataframe
every student's seasonal root meet square error
-----------------------------------
Output:
Figure of long term weekly prediction root mean square errors
"""
# Request the parameters for the plots
y_low = (
input('Please introduce the lower limit for y-Axis (Hit enter for \
default value 0):'))
y_max = (
input('Please introduce the upper limit for y-Axis (Hit enter for \
default values):'))
column_weeks = [i for i in seasonal_rmse.columns]
# Markers for the plot
markers = [
'o', 'v', '^', 'D', '>', 's', 'P', 'X', '<', '>', 'X', 'o', 'v', 's',
'^', 'P', '<', 'D', 's'
]
# Get the array of firstnames for the plot
firstnames = ef.getFirstNames()
# plotting
fig10, ax = plt.subplots()
ax.plot(seasonal_rmse)
for i, line in enumerate(ax.get_lines()):
line.set_marker(markers[i])
ax.set_xlabel('Weeks', fontsize=13, fontweight='bold')
ax.set_ylabel("Root Mean Square Error", fontsize=13, fontweight='bold')
ax.set_title("Seasonal Root Mean Square Error",
fontsize=13,
fontweight='bold')
# Assigns the limits for y-axis based on user's input
if y_low == '' and y_max != '':
ax.set_ylim(seasonal_rmse[column_weeks].min().min(), float(y_max))
elif y_max == '' and y_low != '':
ax.set_ylim(float(y_low), seasonal_rmse[column_weeks].max().max())
elif y_max == '' and y_low == '':
ax.set_ylim(seasonal_rmse[column_weeks].min().min(),
seasonal_rmse[column_weeks].max().max())
else:
ax.set_ylim(float(y_low), float(y_max))
# showing the legend
ax.legend(firstnames,
loc='lower center',
bbox_to_anchor=(.5, -0.4),
ncol=6)
fig10.patch.set_facecolor('xkcd:white')
plt.show()
fig10.savefig(savepath)
def plot_class_forecasts(df, week_flows, leadtime, type_plot):
""" plot_class_forecasts()
---------------------------------------------------------------------
This function plots the forecasts submitted by each student for both
Week 1 & 2 Forecasts. In addition, is capable of plotting the absolute
error in regards the observed value.
---------------------------------------------------------------------
Parameters:
df = Dataframe
Includes the weekly forecast values for Week 1 and 2 for each student.
week_flows = Dataframe
Observed flows per week obtained from USGS.
leadtime: int
leadtime for the forecast. It can only be 1 or 2
type_plot: string
Enter 'forecasts' to plot all submitted values, or 'abs_error'
to plot the deviation from the observed value.
---------------------------------------------------------------------
Outputs: Plot of the forecasted values or the absolute error depending the
user input
"""
# Request the parameters for the plots
y_low = (
input('Please introduce the lower limit for y-Axis (Hit enter for \
default value 0):'))
y_max = (
input('Please introduce the upper limit for y-Axis (Hit enter for \
default values):'))
plot_weeks_inp = input('Please introduce the list of weeks to consider as \
["Week #", "Week #", ...]. Otherwise, if you want to include all weeks\
hit enter:')
if plot_weeks_inp == '':
column_weeks = [i for i in df.columns]
else:
column_weeks = [i for i in df.columns if i in plot_weeks_inp]
# Markers for the plot
markers = [
'o', 'v', '^', 'D', '>', 's', 'P', 'X', '<', '>', 'X', 'o', 'v', 's',
'^', 'P', '<', 'D', 's'
]
# Get the array of firstnames for the plot
firstnames = ef.getFirstNames()
# Trim and set index the same weekly flow (start 8/23)
weekly_flows = week_flows.iloc[1:len(column_weeks) + 1, 3:4]
weekly_flows.set_index(df.columns, append=False, inplace=True)
# Assign values depending the plot type selected
if type_plot == 'abs_error':
df = df.T.subtract(weekly_flows['observed'], axis=0).T
plot_ylabel = "Deviation from Weekly Avg Flow [cfs]"
plot_title = 'Absolute Error in ' + str(
leadtime) + ' Week Forecast for \n\
HAS-Tools Class'
elif type_plot == 'forecast':
plot_ylabel = "Weekly Avg Flow [cfs]"
plot_title = str(leadtime) + ' Week Forecast for HAS-Tools Class \n '
# Plotting process
fig, ax = plt.subplots()
ax.plot(df.T)
for i, line in enumerate(ax.get_lines()):
line.set_marker(markers[i])
# Plot observed flow if the selected plot is the forecast
if type_plot == 'forecast':
ax.plot(column_weeks,
weekly_flows['observed'],
color='black',
marker='o',
linestyle='--',
linewidth=3)
plot_labels = firstnames + ['Observed Flow']
elif type_plot == 'abs_error':
plot_labels = firstnames
# Format for labels and plot title
ax.set_xlabel('Weeks \n', fontsize=13, fontweight='bold')
ax.set_ylabel(plot_ylabel, fontsize=13, fontweight='bold')
ax.set_title(plot_title, fontsize=15, fontweight='bold')
# Assigns the limits for y-axis based on user's input
if y_low == '' and y_max != '':
ax.set_ylim(df[column_weeks].min().min(), float(y_max))
elif y_max == '' and y_low != '':
ax.set_ylim(float(y_low), df[column_weeks].max().max())
elif y_max == '' and y_low == '':
ax.set_ylim(df[column_weeks].min().min(), df[column_weeks].max().max())
else:
ax.set_ylim(float(y_low), float(y_max))
ax.legend(plot_labels,
loc='lower center',
bbox_to_anchor=(.5, -0.4),
ncol=6)
fig.set_size_inches(9, 5)
fig.patch.set_facecolor('xkcd:white')
plt.show()