def f1():
df = pd.merge(
df_review.groupby("userId")["reviewLength"].mean().reset_index(),
df_reviewer,
on="userId")
plotBucket(df,
"numBids",
"reviewLength",
x_label="Number of Bids",
y_label="Average Review Length",
x_percentile=False,
title="Review Quality vs. Number of Bids",
numBuckets=7,
xlim=[0, 100])
plotFrequencyHistogram(df,
'numBids',
'Number of Bids',
myBins=[0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50])
def f1():
df = pd.merge(
df_review.groupby("userId")["reviewLength"].mean().reset_index(),
df_reviewer,
on="userId")
plotBucket(
df,
"numBids",
"reviewLength",
x_label="Number of Bids",
y_label="Average Review Length",
x_percentile=False,
title="Review Quality vs. Number of Bids",
numBuckets=7,
xlim=[0,100]
)
plotFrequencyHistogram(
df,
'numBids',
'Number of Bids',
myBins=[0,5,10,15,20,25,30, 35, 40, 45, 50]
)
'4/12',
'4/13',
'4/14',
'4/15',
'4/16',
'4/17',
]
dates_5Days = transformDates(labels_5Days)
datesMidDay = [
.5 * (dates_5Days[i] + dates_5Days[i + 1])
for i in range(len(dates_5Days) - 1)
]
plotFrequencyHistogram(df,
'time',
'Submission Date',
myBins=dates_5Days,
plotMean=False)
xticks(datesMidDay, labels_5Days[:-1])
plotDeadline()
#Plot 2 -- Submissions in the last 12 hours
labels_12Hours = [
'4/14 12:00 PM',
'4/14 2:00 PM',
'4/14 4:00 AM',
'4/14 6:00 PM',
'4/14 8:00 PM',
'4/14 10:00 PM',
'4/15 12:00 AM',
'4/15 2:00 AM',
more_scramble = set()
for id, reviewer in loader.reviewers.iteritems():
revs = reviewer.reviews
if len(revs) > 7:
sorted_rev = sorted(revs, key=lambda x: int(x.time.strftime('%s')))
paperIds = [r.paper.id for r in sorted_rev]
numInv = countInversions(paperIds)
inversions[id] = numInv
if numInv > 15:
more_scramble.add(id)
else:
less_scramble.add(id)
df_order = pd.DataFrame(inversions.values())
df_order.columns = ['order_stat']
plotFrequencyHistogram(df_order,
'order_stat',
"# perm inversions",
myBins=np.linspace(0, 50, 10))
df_paper = pd.read_pickle("savedFrames/iteration5/paperTable")
df_review = pd.read_pickle("savedFrames/iteration5/reviewTable")
df = pd.merge(df_review, df_paper, on="paperId")
df['time'] = df['time'].values.astype(datetime.datetime)
df["agree"] = (df["rating"] > 0) == df["accepted"]
df["positive"] = df["rating"] > 0
df["absRating"] = df["rating"].abs()
df_less = df[df['userId'].isin(less_scramble)]
df_more = df[df['userId'].isin(more_scramble)]
def transformDates(dateLabels):
return list(
)
plot(
[0, 0],
plt.ylim(),
color='red',
linewidth=.5,
linestyle="--")
legend(
[p1, p2, p3],
["High Similarity", "Medium Similarity", "Low Similarity"],
loc=3)
#Basic Frequency Plots
plotFrequencyHistogram(
df_rating,
maxSimCol,
"Most Experienced Author / Reviewer Similarity",
color="#62FFBB",
myBins=None)
plotFrequencyHistogram(
df_paper,
"avgRating",
"Average Rating for a Paper",
color="#B9E84D",
myBins=np.arange(-3, 3.5, .5))
plotFrequencyHistogram(
df_paper,
"authorsMaxPastPaper",
"Past Paper Count of Most Experienced Author",
color="#E8634D",
'4/11',
'4/12',
'4/13',
'4/14',
'4/15',
'4/16',
'4/17',
]
dates_5Days = transformDates(labels_5Days)
datesMidDay = [.5*(dates_5Days[i] + dates_5Days[i+1])
for i in range(len(dates_5Days) - 1)]
plotFrequencyHistogram(
df,
'time',
'Submission Date',
myBins=dates_5Days,
plotMean=False
)
xticks(datesMidDay, labels_5Days[:-1])
plotDeadline()
#Plot 2 -- Submissions in the last 12 hours
labels_12Hours = [
'4/14 12:00 PM',
'4/14 2:00 PM',
'4/14 4:00 AM',
'4/14 6:00 PM',
'4/14 8:00 PM',
'4/14 10:00 PM',
'4/15 12:00 AM',
x_label="Similarity",
y_label="Rating",
delta=15,
title=
"Average Rating vs. Similarity of Reviewer to Most Experienced Author",
marker="s",
color="green",
xlim=[0, 100]
)
setp(p, linewidth=2, alpha=1)
#Basic Frequency Plots
plotFrequencyHistogram(
df,
"rating",
"Paper Ratings",
color="#B9E84D",
myBins=[-3.6, -2.4, -1.2, 0, 1.2, 2.4, 3.6],
plotMean=False)
xticks([-3, -1.8, -.6, .6, 1.8, 3],
["Strong Reject", "Reject", "Weak Reject",
"Weak Accept", "Accept", "Strong Accept"])
colors =\
["#FF1C00", "#FF5C54", "#FFA6A1", "#A1D3FF", "#54B2FF", "#006FFF"]
for container in plt.gca().containers:
for i, child in enumerate(container.get_children()):
child.set_color(colors[i])
plotFrequencyHistogram(
df_reviewer[df_reviewer["numReviews"] >= 8],