def create_no_open_issues_dev(df, sys_name):
compute_no_open_issues(df)
fig, axs = plt.subplots(figsize=(6, 3))
x_col = "created"
y_col = "no_open_iss"
df.plot.scatter(x=x_col, y=y_col, s=1, ax=axs, rot=0)
axs.set_ylabel("Number of Open Issues")
axs.set_xlabel("Ticket Creation Time")
slope, intercept, xf_dates, yf = compute_lin_reg(df, x_col, y_col)
axs.plot(xf_dates, yf, c="orange")
fig.savefig(
f"data/output/{sys_name[:3]}_no_open_issues.png", bbox_inches="tight"
)
# This code is only to generate the values used in the paper's text
if sys_name == "cassandra":
dt_a = "2010-01-01"
elif sys_name == "gaffer":
dt_a = "2016-01-01"
dt_b = "2021-01-01"
ya = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_a)]))[0] + intercept
yb = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_b)]))[0] + intercept
print(
f"increase in open_iss/year: {slope * (10 ** 9) * 60 * 60 * 24 * 365}",
f"no open issues {dt_a}: {ya}",
f"no open issues {dt_b}: {yb}",
)
return fig, slope, intercept
def compute_lead_t_dev(df, sys_name):
df_small = filter_iqr_outliers(df, "t_lead_s")
df_small["t_lead_h"] = df_small.t_lead_s / 60 / 60 # Convert it to hours
fig, axs = plt.subplots(figsize=(6, 3))
x_col = "resolved"
y_col = "t_lead_h"
df_small.plot.scatter(x=x_col, y=y_col, s=1, ax=axs, rot=0)
axs.set_ylabel("$t_{lead}$ in hours")
axs.set_xlabel("Ticket Close Time")
slope, intercept, xf_dates, yf = compute_lin_reg(df_small, x_col, y_col)
axs.plot(xf_dates, yf, c="orange")
fig.savefig(f"data/output/{sys_name[:3]}_lead_t.png", bbox_inches="tight")
# This code is only to generate the values used in the paper's text
if sys_name == "cassandra":
dt_a = "2010-01-01"
elif sys_name == "gaffer":
dt_a = "2016-01-01"
dt_b = "2021-01-01"
ya = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_a)]))[0] + intercept
yb = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_b)]))[0] + intercept
print(
f"increase in h/year: {slope * (10 ** 9) * 60 * 60 * 24 * 365}",
f"closing time {dt_a}: {ya}",
f"closing time {dt_b}: {yb}",
)
return fig, slope
def create_no_contributors_dev(sys_name):
df = compute_no_contributors(sys_name)
fig, axs = plt.subplots(figsize=(6, 3))
x_col = "release_date"
y_col = "no_contributors"
# df.plot.scatter(x=x_col, y=y_col, s=1, ax=axs, rot=45)
df.plot(
x=x_col,
y=y_col,
ax=axs,
linestyle=":",
marker="o",
rot=45,
legend=False,
)
axs.set_ylabel("Number of Contributors")
axs.set_xlabel("Release Date")
slope, intercept, xf_dates, yf = compute_lin_reg(df, x_col, y_col)
axs.plot(xf_dates, yf, c="orange")
fig.savefig(
f"data/output/{sys_name[:3]}_no_contributors.png", bbox_inches="tight"
)
# This code is only to generate the values used in the paper's text
if sys_name == "cassandra":
dt_a = "2010-01-01"
elif sys_name == "gaffer":
dt_a = "2016-01-01"
dt_b = "2021-01-01"
ya = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_a)]))[0] + intercept
yb = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_b)]))[0] + intercept
print(
f"increase in contr/year: {slope * (10 ** 9) * 60 * 60 * 24 * 365}",
f"no contributors {dt_a}: {ya}",
f"no contributors {dt_b}: {yb}",
)
return fig, slope, intercept
def compute_contribcompl_dev2(df, sys_name):
fig, axs = plt.subplots(figsize=(6, 3))
x_col = "month"
y_col = "avg_contrib_compl"
df.plot.scatter(x=x_col, y=y_col, s=1, ax=axs, rot=0)
axs.set_ylabel("Avg. Contribution Complexity per Month")
axs.set_xlabel("Resolved Date")
slope, intercept, xf_dates, yf = compute_lin_reg(df, x_col, y_col)
axs.plot(xf_dates, yf, c="orange")
axs.set_yticks(list(range(1, 6)))
axs.set_yticklabels(
["low", "moderate", "medium", "elevate", "high"], rotation=90
)
print("Increase in ? per ?", slope)
fig.tight_layout()
# fig.savefig(
# f"data/output/{sys_name[:3]}_contribcompl.png", bbox_inches="tight"
# )
return fig, slope
def compute_contribcompl_dev(df, sys_name):
fig, axs = plt.subplots(figsize=(6, 3))
x_col = "resolved"
y_col = "contrib_complexity"
df.plot.scatter(x=x_col, y=y_col, s=1, ax=axs, rot=0)
axs.set_ylabel("Contribution Complexity")
axs.set_xlabel("Resolved Date")
slope, intercept, xf_dates, yf = compute_lin_reg(df, x_col, y_col)
axs.plot(xf_dates, yf, c="orange")
axs.set_yticks(list(range(1, 6)))
axs.set_yticklabels(
["low", "moderate", "medium", "elevate", "high"], rotation=45
)
# This code is only to generate the values used in the paper's text
if sys_name == "cassandra":
dt_a = "2010-01-01"
elif sys_name == "gaffer":
dt_a = "2016-01-01"
dt_b = "2021-01-01"
ya = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_a)]))[0] + intercept
yb = slope * pd.to_numeric(pd.Series([pd.to_datetime(dt_b)]))[0] + intercept
print(
f"increase in cc/year: {slope * (10 ** 9) * 60 * 60 * 24 * 365}",
f"cc {dt_a}: {ya}",
f"cc {dt_b}: {yb}",
)
fig.tight_layout()
fig.savefig(
f"data/output/{sys_name[:3]}_contribcompl.png", bbox_inches="tight"
)
return fig, slope