def actualPred(y_true, y_pred):
"""Plot actual vs predicted line plots
Parameters
----------
y_true : Series or vector
The ground truth
y_pred : Series or vector
Predicted values
"""
datum = pd.DataFrame({
"date": range(y_true.shape[0]),
"Actual": y_true,
"Prediction": y_pred
})
datum = pd.melt(datum,
id_vars=['date'],
value_vars=['Actual', 'Prediction'])
p = (
ggplot(datum, aes(x='date')) +
geom_line(aes(y='value', color='variable')) # line plot
+ labs(x='date', y='Solar Output') + plotnine.theme_538() +
plotnine.theme(figure_size=(10, 6)))
print(p)
def plot_score(df, plot_fn):
f = (p9.ggplot(df, p9.aes(x="emotion_cat", y="score")) +
p9.geom_boxplot() + p9.labs(x="Model", y="EMOTION FEEL Score") +
p9.theme_538() + p9.theme(legend_position="top",
legend_direction="horizontal",
figure_size=(10, 5)) +
p9.theme(plot_background=p9.element_rect(
fill=BG_COLOR, color=BG_COLOR, size=1)))
f.save(plot_fn)
def scatterplot(cls, df):
Utils.check_and_make_dir("Figures/Scatterplots")
df = df[(df['index'] != 'Overall') &
(df['index'] != 'No ROI')] # Remove No ROI and Overall rows
df = df.groupby([config.table_cols, config.table_rows]).apply(
lambda x: x.sort_values(['Mean'])) # Group by parameters and sort
df = df.reset_index(drop=True) # Reset index to remove grouping
scatterplots = ['roi_ordered', 'stat_ordered']
if config.table_row_order == 'roi':
scatterplots.remove('stat')
elif config.table_row_order == 'statorder':
scatterplots.remove('roi_ordered')
for scatterplot in scatterplots:
if config.verbose:
print(f"Saving {scatterplot} scatterplot!")
if scatterplot == 'roi_ordered':
roi_ord = pd.Categorical(df['index'],
categories=df['index'].unique()
) # Order rows based on first facet
else:
roi_ord = pd.Categorical(
df.groupby(['MB', 'SENSE'
]).cumcount()) # Order each facet individually
figure_table = (
pltn.ggplot(df, pltn.aes(x="Mean", y=roi_ord)) +
pltn.geom_point(na_rm=True, size=1) + pltn.geom_errorbarh(
pltn.aes(xmin="Mean-Conf_Int_95", xmax="Mean+Conf_Int_95"),
na_rm=True,
height=None) + pltn.xlim(0, None) +
pltn.scale_y_discrete(labels=[]) +
pltn.ylab(config.table_y_label) +
pltn.xlab(config.table_x_label) +
pltn.facet_grid('{rows}~{cols}'.format(rows=config.table_rows,
cols=config.table_cols),
drop=True,
labeller="label_both") +
pltn.theme_538() # Set theme
+ pltn.theme(
panel_grid_major_y=pltn.themes.element_line(alpha=0),
panel_grid_major_x=pltn.themes.element_line(alpha=1),
panel_background=pltn.element_rect(fill="gray", alpha=0.1),
dpi=config.plot_dpi))
figure_table.save(
f"Figures/Scatterplots/{scatterplot}_scatterplot.png",
height=config.plot_scale,
width=config.plot_scale * 3,
verbose=False,
limitsize=False)
def plot_rank_full(df, plot_fn):
f = (p9.ggplot(df, p9.aes(x="emotion_cat", y="ratio", fill="factor(rank)"))
+ p9.geom_bar(stat="identity") + p9.facet_wrap("cluster_labels_6") +
p9.labs(x="Model", y="Proportion (%)", fill="Rank") + p9.theme_538() +
p9.theme(legend_position="top",
legend_direction="horizontal",
figure_size=(10, 5)) +
p9.theme(plot_background=p9.element_rect(
fill=BG_COLOR, color=BG_COLOR, size=1),
axis_text_x=p9.element_text(rotation=45, hjust=1)))
f.save(plot_fn)
def scatter(y_true, y_pred):
"""Plot actual vs predicted scatterplot
Parameters
----------
y_true : Series or vector
The ground truth
y_pred : Series or vector
Predicted values
"""
datum = pd.DataFrame({"Actual": y_true, "Prediction": y_pred})
p = (
ggplot(datum, aes(x='Actual', y="Prediction", color='"#9B59B6"')) +
geom_point() # line plot
+ labs(x='Actual', y='Prediction') + plotnine.theme_538() +
plotnine.theme(figure_size=(10, 6)))
print(p)
def get_lag_corr(y_actual, y_pred, num_lags):
"""Calculates & plots Lag Correlation
Parameters
----------
y_actual : Series or vector
The ground truth
y_pred : Series or vector
Predicted values
num_lags : int
Lag to consider - range (0, num_lags)
"""
lags = []
for c in range(num_lags):
lagged = pd.Series(y_pred).shift(c)
lags.append(
scipy.stats.spearmanr(lagged, y_actual, nan_policy='omit')[0])
datum = pd.DataFrame({"Lags": range(len(lags)), "Lag-Coefficient": lags})
p = (ggplot(datum, aes(x='Lags')) + geom_line(aes(y='Lag-Coefficient')) +
labs(x='Lag', y='Coefficient') + plotnine.theme_538() +
plotnine.theme(figure_size=(10, 6)))
print(p)
def test_theme_538(self):
p = self.g + labs(title='Theme 538') + theme_538()
assert p + _theme == 'theme_538'
def histogram_make(roi, combined_raw_df, list_rois, config, xlimit,
save_function, find_xlim_function):
if combined_raw_df.empty:
if config.verbose:
print(
'INFO: Histograms cannot be made for the No ROI category.')
return
else:
thisroi = list_rois[roi]
figure = (
pltn.ggplot(combined_raw_df, pltn.aes(x="voxel_value")) +
pltn.theme_538() + pltn.geom_histogram(
binwidth=config.histogram_binwidth,
fill=config.histogram_fig_colour,
boundary=0,
na_rm=True
) # Boundary centers the bars, na_rm cancels error from setting an xlimit
+ pltn.facet_grid(
f"{config.histogram_fig_y_facet}~{config.histogram_fig_x_facet}",
drop=True,
labeller="label_both") +
pltn.labs(x=config.histogram_fig_label_x,
y=config.histogram_fig_label_y) +
pltn.theme(
panel_grid_minor_x=pltn.themes.element_line(alpha=0),
panel_grid_major_x=pltn.themes.element_line(alpha=1),
panel_grid_major_y=pltn.element_line(alpha=0),
plot_background=pltn.element_rect(fill="white"),
panel_background=pltn.element_rect(fill="gray", alpha=0.1),
axis_title_x=pltn.element_text(
weight='bold', color='black', size=20),
axis_title_y=pltn.element_text(
weight='bold', color='black', size=20),
strip_text_x=pltn.element_text(
weight='bold', size=10, color='black'),
strip_text_y=pltn.element_text(
weight='bold', size=10, color='black'),
axis_text_x=pltn.element_text(size=10, color='black'),
axis_text_y=pltn.element_text(size=10, color='black'),
dpi=config.plot_dpi))
# Display mean or median as vertical lines on plot
if config.histogram_show_mean or config.histogram_show_median:
figure += pltn.geom_vline(pltn.aes(xintercept="stat_value",
color="Statistic"),
size=config.histogram_stat_line_size)
figure += pltn.scale_color_manual(values=[
config.colorblind_friendly_plot_colours[3],
config.colorblind_friendly_plot_colours[1]
])
# Display legend for mean and median
if not config.histogram_show_legend:
figure += pltn.theme(legend_position='none')
if xlimit:
# Set y limit of figure (used to make it the same for every barchart)
figure += pltn.xlim(-1, xlimit)
thisroi += '_same_xlim'
else:
figure += pltn.xlim(-1, None)
returned_xlim = 0
if config.use_same_axis_limits in ('Same limits',
'Create both') and xlimit == 0:
returned_xlim = find_xlim_function(thisroi, figure, 'xaxis')
if config.use_same_axis_limits == 'Same limits' and xlimit == 0:
return returned_xlim
elif xlimit != 0:
folder = 'Same_xaxis'
else:
folder = 'Different_xaxis'
# Suppress Pandas warning about alignment of non-concatenation axis
warnings.simplefilter(action='ignore', category=FutureWarning)
save_function(figure, thisroi, config, folder, 'histogram')
warnings.simplefilter(action='default', category=FutureWarning)
return returned_xlim
def barchart_make(roi, df, list_rois, config, ylimit, save_function,
find_ylim_function):
thisroi = list_rois[roi]
current_df = df.loc[df['index'] == thisroi]
current_df = current_df.sort_values([config.single_roi_fig_x_axis])
current_df = current_df.reset_index(
drop=True) # Reset index to remove grouping
current_df[config.single_roi_fig_x_axis] = pd.Categorical(
current_df[config.single_roi_fig_x_axis],
categories=current_df[config.single_roi_fig_x_axis].unique())
figure = (
pltn.ggplot(
current_df,
pltn.aes(x=config.single_roi_fig_x_axis,
y='Mean',
ymin="Mean-Conf_Int_95",
ymax="Mean+Conf_Int_95",
fill='factor({colour})'.format(
colour=config.single_roi_fig_colour))) +
pltn.theme_538() + pltn.geom_col(position=pltn.position_dodge(
preserve='single', width=0.8),
width=0.8,
na_rm=True) +
pltn.geom_errorbar(size=1,
position=pltn.position_dodge(
preserve='single', width=0.8)) +
pltn.labs(x=config.single_roi_fig_label_x,
y=config.single_roi_fig_label_y,
fill=config.single_roi_fig_label_fill) +
pltn.scale_x_discrete(labels=[]) +
pltn.theme(panel_grid_major_x=pltn.element_line(alpha=0),
axis_title_x=pltn.element_text(
weight='bold', color='black', size=20),
axis_title_y=pltn.element_text(
weight='bold', color='black', size=20),
axis_text_y=pltn.element_text(size=20, color='black'),
legend_title=pltn.element_text(size=20, color='black'),
legend_text=pltn.element_text(size=18, color='black'),
subplots_adjust={'right': 0.85},
legend_position=(0.9, 0.8),
dpi=config.plot_dpi) +
pltn.geom_text(pltn.aes(y=-.7, label=config.single_roi_fig_x_axis),
color='black',
size=20,
va='top') + pltn.scale_fill_manual(
values=config.colorblind_friendly_plot_colours))
if ylimit:
# Set y limit of figure (used to make it the same for every barchart)
figure += pltn.ylim(None, ylimit)
thisroi += '_same_ylim'
returned_ylim = 0
if config.use_same_axis_limits in ('Same limits',
'Create both') and ylimit == 0:
returned_ylim = find_ylim_function(thisroi, figure, 'yaxis')
if config.use_same_axis_limits == 'Same limits' and ylimit == 0:
return returned_ylim
elif ylimit != 0:
folder = 'Same_yaxis'
else:
folder = 'Different_yaxis'
save_function(figure, thisroi, config, folder, 'barchart')
return returned_ylim
def test_theme_538(self):
p = self.g + labs(title='Theme 538') + theme_538()
assert p + _theme == 'theme_538'
def plot_performance(df, report_year, eval_period):
"""
Plot metric-specific performance for a set of stocks over time. Reference:
https://www.buffettsbooks.com/how-to-invest-in-stocks/intermediate-course/lesson-20/
:param df: DataFrame containing stock tickers and the columns specified below
:param report_year: Year of most recent financial report
:param eval_period: Number of years prior to most recent report to be analyzed
:return: A list of ggplot objects
:rtype: List
"""
start_year = report_year - eval_period
df = df.loc[df['year'] >= start_year]
df = df[[
'symbol', 'year', 'eps', 'bookValuePerShare', 'roe', 'currentRatio',
'debtToEquity'
]]
df['roe'] = df['roe'].apply(lambda x: x * 100.0)
df = df.rename(
{
'eps': 'Earnings per Share',
'roe': 'Return on Equity',
'currentRatio': 'Current Ratio',
'debtToEquity': 'Debt to Equity Ratio',
'bookValuePerShare': 'Book Value per Share'
},
axis='columns')
df.sort_values(by=['symbol', 'year'], inplace=True, ascending=True)
df.dropna(inplace=True)
# Commenting out for now, API no longer returning this col in income-statement response
label_dict = {
'Earnings per Share':
'The EPS shows the company\'s profit per share. This chart '
'should have a positive slope over time. Stable results '
'here are extremely important for forecasting future cash '
'flows. Note: if the company\'s book value has increased '
'over time, the EPS should demonstrate similar growth.',
# 'Dividend per Share': 'This chart shows the dividend history of the company. '
# 'This should have a flat to positive slope over time. If '
# 'you see a drastic drop, it may represent a stock split '
# 'for the company. Note: the dividend is taken from a '
# 'portion of the EPS, the remainder goes to the book value.',
'Book Value per Share':
'The book value represents the liquidation value of the '
'entire company (per share). It\'s important to see '
'this number increasing over time. If the company pays a'
' high dividend, the book value may grow at a slower '
'rate. If the company pays no dividend, the book value '
'should grow with the EPS each year.',
'Return on Equity':
'Return on equity is very important because it show the '
'return that management has received for reinvesting the '
'profits of the company. If using an intrinsic value '
'calculator, it\'s very important that this number is flat or'
' increasing for accurate results. Find companies with a '
'consistent ROE above 8%.',
'Current Ratio':
'The current ratio helps measure the health of the company in '
'the short term. As a rule of thumb, the current ratio should be'
' above 1.0. A safe current ratio is typically above 1.5. Look '
'for stability trends within the current ratio to see how the '
'company manages their short term risk.',
'Debt to Equity Ratio':
'The debt to equity ratio helps measure the health of '
'the company in the long term. As a rule of thumb, the '
'debt to equity ratio should be lower than 0.5. Look for '
'stability trends within the debt/equity ratio to see how'
' the company manages their long term risk.'
}
wrapper = textwrap.TextWrapper(width=120)
for key, value in label_dict.items():
label_dict[key] = wrapper.fill(text=value)
plots = []
cols = df.columns[2:].tolist()
for metric in cols:
p = (ggplot(df, aes('year', metric, color='symbol')) +
geom_line(size=1, alpha=0.8) + geom_point(size=3, alpha=0.8) +
labs(title=metric, x='Report Year', y='', color='Ticker') +
theme_538() + theme(legend_position='left',
plot_title=element_text(weight='bold')) +
scale_x_continuous(breaks=range(min(df['year']),
max(df['year']) + 1, 1)) +
scale_y_continuous(
breaks=range(min(df[metric].astype(int)),
max(round(df[metric]).astype(int)) + 2, 1)) +
annotate(geom='label',
x=statistics.mean((df['year'])),
y=max(round(df[metric]).astype(int) + 1),
label=label_dict[metric],
size=8,
label_padding=0.8,
fill='#F7F7F7'))
plots.append(p)
return plots
