def venn2_plot(set1 = set(),
set2 = set(),
lab_set1 = 'Set1',
lab_set2 = 'Set2',
linewidth = 1,
color_line = 'black',
alpha_sets = 0.3,
font_sets = False, # False o 'bold'
size_vals_sets = 12,
alpha_inter = 0.3,
font_inter = False, # False o 'bold'
size_vals_inter = 12,
size_label = 12,
font_label = False): # False o 'bold'
v = venn2_unweighted(subsets = (set1, set2), set_labels = (lab_set1, lab_set2))
c = venn2_circles(subsets = (1, 1, 1),
linestyle='--', linewidth= linewidth, color=color_line)
v.get_patch_by_id('10').set_alpha(0)
partes = ['10', '01', '11']
partes2 = ['10', '01']
venn_info = [[i, j] for i, j in zip(v.subset_labels, partes)]
for i in venn_info:
if i[0] != None:
if i[1] in partes2:
v.get_patch_by_id(i[1]).set_alpha(alpha_sets) # i[1] = el conjunto creado, 0 = alpha del conjunto
v.get_label_by_id(i[1]).set_fontweight(font_sets)
v.get_label_by_id(i[1]).set_fontsize(size_vals_sets)
if i[1] == '11': # configurar la intersección independientemente '111'
v.get_patch_by_id('11').set_alpha(alpha_inter) # i[1] = el conjunto creado, 0 = alpha del conjunto
v.get_label_by_id('11').set_fontweight(font_inter)
v.get_label_by_id('11').set_fontsize(size_vals_inter)
for text in v.set_labels:
text.set_fontsize(size_label)
text.set_fontweight(font_label)
def init_compare(input1, input2, labels=None):
input1_df = df.from_csv(input1, index_col=False)
input2_df = df.from_csv(input2, index_col=False)
input1_df_index = [
';'.join(_i.split('\t')[:5]) for _i in list(input1_df.Otherinfo)
]
input2_df_index = [
';'.join(_i.split('\t')[:5]) for _i in list(input2_df.Otherinfo)
]
if labels:
ax = venn2_unweighted(
[set(input1_df_index), set(input2_df_index)], set_labels=labels)
return ax
else:
input1_df.index = input1_df_index
input2_df.index = input2_df_index
result = {}
result['small_one_unique'] = input1_df.loc[
set(input1_df_index).difference(set(input2_df_index)), :]
result['big_one_should_in_small'] = input2_df.loc[
gene_filter(input2_df, _in=gene_list), :]
result['shared'] = input1_df.loc[
set(input1_df_index).intersection(set(input2_df_index)), :]
return result
def process_venn(out_dir, venn):
from matplotlib_venn import venn3, venn3_unweighted, venn2_unweighted
plt.clf()
pvalue_cutoff = venn["pvalue_cutoff"]
biofluid_regions = venn["biofluid_regions"]
disorders = venn["disorders"]
genes = {}
for biofluid_region in biofluid_regions:
col = 0
for disorder in disorders:
filename = "data/out/" + biofluid_region + "/" + disorder + "/lrt.tsv"
if os.path.exists(filename):
df = pd.read_csv(filename, sep='\t', index_col=0)
# Filter genes with pvalue less than cutoff
df = df[df["pvalue"] < pvalue_cutoff]
# Add to list
if disorder in genes:
genes[disorder] = genes[disorder] + df.index.tolist()
else:
genes[disorder] = df.index.tolist()
else:
genes[disorder] = []
# Find unique genes per disorder
for disorder in disorders:
genes[disorder] = set(genes[disorder])
a = genes[disorders[0]]
b = genes[disorders[1]]
intersection = a & b
a_only = a - intersection
b_only = b - intersection
fig = venn2_unweighted(subsets=(len(a - (a & b)), len(b - (a & b)),
len((a & b))),
set_labels=tuple(disorders),
alpha=0.5)
plt.text(x=0.05,
y=-0.2,
s="\n".join(list(intersection)),
color='black',
bbox=dict(facecolor='orange', alpha=0.5))
plt.text(x=-0.84,
y=-0.25,
s="\n".join(list(a_only)),
color='black',
bbox=dict(facecolor='red', alpha=0.5))
plt.text(x=0.66,
y=-0.2,
s="\n".join(list(b_only)),
color='black',
bbox=dict(facecolor='green', alpha=0.5))
plt.title(venn["title"])
plt.savefig(out_dir + "/venn.png")
return
def motifplot(G, proatom_order, lig_name, MotifFolder):
'''this function is to make a picture combined bar plot with pie chart characterizing the motif biochemical properties,
and make a venn plot to show overlaps of protein pockets containing different motifs for one ligand
o input: (1) G: the number of function groups of the ligand , type: int
(2) proatom_order: a dict of different binding motifs for a ligand
(3) lig_name: the ligand name
(4) MotifFolder: the location of a folder to place plots that display motif features
o output: save the pictures into motifFolder folder.
for each of 233 ligands:
return 2 pictures, one is a picture comprising bar plot and pie chart,and the other one is a venn plot.
note: if the number of function groups for one ligand is more than 3, there will be only one picture.
'''
fig, axes = plt.subplots(G, 2, figsize=(8, 3 * G), dpi=300)
fig1, axes1 = plt.subplots(1, 1)
atomorder = sorted(proatom_order.items())
cluster_set = []
labels = []
# plot bar and pie chart displaying motif amino acids distribution and atom properties distribution respectively
for cluster in atomorder:
A, aminoAcids, props = GetMotif(cluster[1])
PlotBarChart(axes[atomorder.index(cluster), 0], aminoAcids)
PlotPieChart(axes[atomorder.index(cluster), 1], props)
cluster_set.append(set(A))
labels.append(cluster[0])
fig.savefig(os.path.join(MotifFolder, "%s_analysis.png" % lig_name),
dpi=300,
bbox_inches="tight")
# make a venn plot to show overlaps of protein pockets containing 2 motifs for one ligand
if G == 2:
venn2_unweighted(cluster_set, labels, ax=axes1)
fig1.savefig(os.path.join(MotifFolder, "%s_provenn.png" % lig_name),
dpi=300)
# make a venn plot to show overlaps of protein pockets containing 3 motifs for one ligand
elif G == 3:
venn3_unweighted(cluster_set, labels, ax=axes1)
fig1.savefig(os.path.join(MotifFolder, "%s_provenn.png" % lig_name),
dpi=300)
def test_pr_28():
import matplotlib_venn as mv
v = mv.venn3((1, 2, 3, 4, 5, 6, 7), subset_label_formatter = None)
assert v.get_label_by_id('010').get_text() == '2'
v = mv.venn3((1, 2, 3, 4, 5, 6, 7), subset_label_formatter = lambda x: 'Value: %+0.3f' % (x / 100.0))
assert v.get_label_by_id('010').get_text() == 'Value: +0.020'
v = mv.venn2((1, 2, 3), subset_label_formatter = None)
assert v.get_label_by_id('01').get_text() == '2'
v = mv.venn2((1, 2, 3), subset_label_formatter = lambda x: 'Value: %+0.3f' % (x / 100.0))
assert v.get_label_by_id('01').get_text() == 'Value: +0.020'
v = mv.venn3_unweighted((1, 2, 3, 4, 5, 6, 7), subset_label_formatter = lambda x: 'Value: %+0.3f' % (x / 100.0))
assert v.get_label_by_id('010').get_text() == 'Value: +0.020'
v = mv.venn2_unweighted((1, 2, 3), subset_label_formatter = lambda x: 'Value: %+0.3f' % (x / 100.0))
assert v.get_label_by_id('01').get_text() == 'Value: +0.020'
def venn(self, amplicons, primers, mismatch):
fig, ax = plt.subplots()
n = len(primers)
if n > 2:
primers = primers[:3]
venn_data = venn3_unweighted(map(lambda k: set(amplicons.get(k)), primers), set_labels = primers)
elif n == 2:
venn_data = venn2_unweighted(map(lambda k: set(amplicons.get(k)), primers), set_labels = primers)
else:
return
[ text.set_fontsize(16) for text in venn_data.set_labels ]
fig_fname = '.'.join(['venn_mismatch{}'.format(mismatch),'pdf'])
plt.tight_layout()
plt.savefig(fig_fname, dpi=1000)
plt.close()
plt.clf()
def test_pr_28():
import matplotlib_venn as mv
v = mv.venn3((1, 2, 3, 4, 5, 6, 7), subset_label_formatter=None)
assert v.get_label_by_id('010').get_text() == '2'
v = mv.venn3((1, 2, 3, 4, 5, 6, 7),
subset_label_formatter=lambda x: 'Value: %+0.3f' %
(x / 100.0))
assert v.get_label_by_id('010').get_text() == 'Value: +0.020'
v = mv.venn2((1, 2, 3), subset_label_formatter=None)
assert v.get_label_by_id('01').get_text() == '2'
v = mv.venn2((1, 2, 3),
subset_label_formatter=lambda x: 'Value: %+0.3f' %
(x / 100.0))
assert v.get_label_by_id('01').get_text() == 'Value: +0.020'
v = mv.venn3_unweighted((1, 2, 3, 4, 5, 6, 7),
subset_label_formatter=lambda x: 'Value: %+0.3f' %
(x / 100.0))
assert v.get_label_by_id('010').get_text() == 'Value: +0.020'
v = mv.venn2_unweighted((1, 2, 3),
subset_label_formatter=lambda x: 'Value: %+0.3f' %
(x / 100.0))
assert v.get_label_by_id('01').get_text() == 'Value: +0.020'
### Venn diagram of clonal sequences for each sample
fig = plt.figure()
ax = fig.add_subplot(111)
#colors = {'10X': 'skyblue', 'Immunoseq': 'purple', 'lcRNA-Seq': 'red', 'mRNA-Seq': 'green', '1ng': 'orange'}
colors = defaultdict(lambda: 'white')
abundance1 = int(100. * sharedcount[0] / totals[0])
abundance2 = int(100. * sharedcount[1] / totals[1])
lowest_abun = min([abundance1, abundance2])
intersect_label = "(%s%%/%s%%)" % (abundance1, abundance2)
figure = venn2_unweighted(subsets=(len(set_list[0]) - intersect,
len(set_list[1]) - intersect, intersect),
set_labels=(argv[2].split('.')[0],
argv[3].split('.')[0]),
set_colors=(colors[argv[2].split('.')[0]],
colors[argv[3].split('.')[0]]))
venn2_circles(subsets=(1, 1, 1), linewidth=3)
print(len(set_list[0]) - intersect, len(set_list[1]) - intersect, intersect)
for text in figure.set_labels:
text.set_fontsize(36)
for text in figure.subset_labels:
text.set_fontsize(36)
lbl1 = figure.get_label_by_id("A")
lbl2 = figure.get_label_by_id("B")
x1, y1 = lbl1.get_position()
x2, y2 = lbl2.get_position()
if args.unweighted:
venn3_unweighted(venn_sizes, set_labels=args.labels)
venn3_circles({'100': 1, '010': 1, '001': 1, '110': 1,
'101': 1, '011': 1, '111': 1})
else:
venn3(venn_sizes, set_labels=args.labels)
venn3_circles(venn_sizes)
else:
from matplotlib_venn import venn2, venn2_circles, venn2_unweighted
venn_data = get_venn2_subsets(*[set_info[x] for x in set_info])
venn_sizes = {x:len(venn_data[x]) for x in venn_data}
if args.unweighted:
venn2_unweighted(venn_sizes, set_labels=args.labels)
venn2_circles({'10': 1, '01': 1, '11': 1})
else:
venn2(venn_sizes, set_labels=args.labels)
venn2_circles(venn_sizes)
if args.savefig:
# without bbox_inches, the saved figure has truncated axes.
plt.savefig(args.savefig, bbox_inches='tight')
# print parsed output. fancy lambda basically enforces the order of keys
# being printed out, i.e. '111', '110', '101', '011', '100', '010', '001'.
for v in sorted(venn_data, key=lambda x: (x.count('1'), int(x)), reverse=True):
print (', '.join([args.labels[n] for n, x in enumerate(v) if x == '1']),
end=':\n')
if args.summary:
def vis():
print("Extracting history... ")
# creating .json file
recentsongs = spotify.current_user_recently_played(limit=50)
out_file = open("recentsongs.json", "w")
out_file.write(json.dumps(recentsongs, sort_keys=True, indent=2))
out_file.close()
f = open('recentsongs.json', )
data = json.load(f)
f.close()
track_id = []
track_name = []
album_name = []
artist_name = []
track_time = []
# creating arrays for storing id and name
for i in recentsongs['items']:
track_id.append(i['track']['id'])
track_name.append(i['track']['name'])
album_name.append(i['track']['album']['name'])
artist_name.append(i['track']['artists'][0]['name'])
temp = i['played_at']
track_time.append(re.sub('.[0-9].[000-999]Z', '', temp))
print("Accessing features of all 50 tracks... ")
features = []
tracks = {}
for track in track_id:
features.append(sp.audio_features(track))
# initialising all tracks with corresponding feature values and storing in a dictionary
for i in range(len(track_name)):
tracks[i + 1] = {}
for i in range(len(track_name)):
tracks[i + 1]['time'] = track_time[i]
tracks[i + 1]['name'] = track_name[i]
tracks[i + 1]['id'] = track_id[i]
tracks[i + 1]['album'] = album_name[i]
tracks[i + 1]['artist'] = artist_name[i]
tracks[i + 1]['acousticness'] = features[i][0]['acousticness']
tracks[i + 1]['danceability'] = features[i][0]['danceability']
tracks[i + 1]['energy'] = features[i][0]['energy']
tracks[i + 1]['instrumentalness'] = features[i][0]['instrumentalness']
tracks[i + 1]['liveness'] = features[i][0]['liveness']
tracks[i + 1]['loudness'] = features[i][0]['loudness']
tracks[i + 1]['speechiness'] = features[i][0]['speechiness']
tracks[i + 1]['tempo'] = features[i][0]['tempo']
tracks[i + 1]['valence'] = features[i][0]['valence']
pop = sp.track(track_id[i])
tracks[i + 1]['popularity'] = pop['popularity']
# creating dictionary to convert into dataframe
feature = [
'time', 'name', 'id', 'album', 'artist', 'acousticness',
'danceability', 'energy', 'instrumentalness', 'liveness', 'loudness',
'speechiness', 'tempo', 'valence', 'popularity'
]
dic_df = {}
print("Converting into dataframe...")
time.sleep(2)
# initialising dictionary
for x in feature:
dic_df[x] = []
for j in range(len(track_name)):
for x in feature:
dic_df[x].extend([tracks[j + 1][x]])
# creating dataframe from dictionary
dataframe = pd.DataFrame.from_dict(
dic_df).iloc[::-1] # .drop_duplicates(subset='name')
pd.set_option('display.width', None)
print(dataframe)
valence_vals = dataframe['valence'].tolist()
less_count, more_count, middle_count = 0, 0, 0
for num in valence_vals:
if num >= 0 and num < 0.5:
less_count += 1
elif num >= 0.5 and num < 0.6:
middle_count += 1
else:
more_count += 1
print("Plotting Graphs... ")
venn2_unweighted(subsets=(less_count, more_count, middle_count),
set_labels=('Low Spirit', 'High Spirit'),
set_colors=('navy', 'lime'),
alpha=0.5)
dataframe.plot.line(x='time', y=['danceability', 'energy', 'valence'])
plt.xticks(rotation=90)
plt.show()
if (less_count / (less_count + more_count + middle_count) > 0.5):
val = input(
"We've noticed you've been listening to quite a few sad songs lately.\nWould you like us to suggest you a curated playlist to lift your mood?(y/n)"
)
if val == "y":
print("Evaluating...")
time.sleep(5)
print(
"In your spotify library, you'll find a playlist named 'Mood Booster'. Enjoy!!"
)
sets,
[
"{} (Total {})".format(name, length)
for (name, length) in zip(names, lengths)
],
)
if len(sets) == 2:
try:
from matplotlib_venn import venn2_unweighted
except ImportError:
sys.exit("Requires the Python library matplotlib_venn")
venn2_unweighted(
sets,
[
"{} (Total {})".format(name, length)
for (name, length) in zip(names, lengths)
],
)
# not sure what I am doing here,
# matplotlib_venn does not want to create a single Venn circle
# stick to the old behavior (rpy and Limma) as much as possible
if len(sets) == 1:
try:
from matplotlib_venn import venn2
except ImportError:
sys.exit("Requires the Python library matplotlib_venn")
venn2((sets[0], set()), [set_data[0][2], ""])
plt.title(all_label)
print(bucket_detailsD)
bucketAB = bucket_detailsB.merge(bucket_detailsA,
how="outer",
on="Bucket_Name",
indicator=True)
print(bucketAB)
New_Bucket = bucketAB.loc[bucketAB._merge ==
"right_only"].iloc[:, [0, 4, 5, 6]]
New_Bucket.columns = bucket_header
print(New_Bucket)
v = vpit.venn2_unweighted(subsets={
"10": len(bucket_detailsB.index),
'01': len(New_Bucket.index),
'11': len(bucket_detailsD.index)
},
set_labels=('Before', 'New'),
alpha=0.5)
matplotlib.pyplot.savefig("temp.png")
data_uri = base64.b64encode(open('temp.png', 'rb').read()).decode('utf-8')
venn_image = '<img src="data:image/png;base64,{0}">'.format(data_uri)
HEADER = """
<html>
<head>
<style>
.panel{margin-bottom20px;background-color:#fff;border:3px solid;border-radius;border-color:#1f0f66;}
.panel-body{margin: 30px;}
deetcont = pd.read_csv("edgeR-Genes/genelists/downDeet.csv", header=None)
perm = pd.read_csv("edgeR-Genes/genelists/upPerm.csv", header=None)
permcont = pd.read_csv("edgeR-Genes/genelists/downPerm.csv", header=None)
genes = pd.read_csv("edgeR-Genes/genelists/listofgenes.csv", header=None)
def compare(s1, s2):
c = len(list(set(s1.iloc[:, 0]) & set(s2.iloc[:, 0])))
return (c)
diff = compare(deet, genes)
len(deet)
fig, axes = plt.subplots(2)
v1 = venn2_unweighted(subsets=(len(deet), len(deetcont),
len(genes) - len(deetcont) - len(deet)),
set_labels=("Up Regulated", "Down Regulated"),
ax=axes[0])
v2 = venn2_unweighted(subsets=(len(perm), len(permcont),
len(genes) - len(permcont) - len(perm)),
set_labels=("Up Regulated", "Down Regulated"),
ax=axes[1])
axes[0].set_title("Deet")
axes[1].set_title("Perm")
fig.tight_layout()
fig.savefig("figures/EdgeR/DeetAndPermRegulation.png",
bbox_inches="tight",
dpi=250)
plt.show()
genes = pd.read_csv("GeneReference/listofgenes.csv", header=None)
body = pd.read_csv("GeneReference/Body.csv")
sets,
[
"{} (Total {})".format(name, length)
for (name, length) in zip(names, lengths)
],
)
if len(sets) == 2:
try:
from matplotlib_venn import venn2_unweighted
except ImportError:
sys.exit("Requires the Python library matplotlib_venn")
venn2_unweighted(
sets,
[
"{} (Total {})".format(name, length)
for (name, length) in zip(names, lengths)
],
)
# not sure what I am doing here,
# matplotlib_venn does not want to create a single Venn circle
# stick to the old behavior (rpy and Limma) as much as possible
if len(sets) == 1:
try:
from matplotlib_venn import venn2
except ImportError:
sys.exit("Requires the Python library matplotlib_venn")
venn2((sets[0], set()), [set_data[0][2], ""])
plt.title(all_label)
sub=[] for condition_nb in set_to_title.values(): if condition_nb in venn_diagram: sub.append(venn_diagram[condition_nb]) else: sub.append(0) # print tuple(sub) # avec les 3 fichiers test, resultat attendu #Only in test1 Only in test2 Common test1 test2 Only in test3 Common test1 test3 Common test2 test3 Common test1 test2 test3 #(8, 6, 1, 7, 0, 2, 0) plt.figure(figsize=(14,10)) # first number : width , second number : height if len(liste_sets) == 2: v = venn2_unweighted(subsets = tuple(sub), set_labels = (liste_sets[0], liste_sets[1])) elif len(liste_sets) == 3: v = venn3_unweighted(subsets = tuple(sub), set_labels = (liste_sets[0], liste_sets[1], liste_sets[2])) for text in v.set_labels: # file name size text.set_fontsize(12) for text in v.subset_labels: # numbers inside circles size text.set_fontsize(16) if args.venn_title is not None : plt.title(args.venn_title) # display title #liste_sets_string = ', '.join(liste_sets) #print liste_sets_string #plt.title(liste_sets_string)
genelist_output_location = '../gene_lists'
namingscheme_parts = list(findOutputFilename(WT_filepath))
WT_cutoff_id = namingscheme_parts[0]
WT_zone_id = namingscheme_parts[1]
venn_output_filename = 'fig1_' + WT_zone_id + '_' + WT_cutoff_id + '.png'
WT_expr = set(readGeneList(WT_filepath))
dgt_expr = set(readGeneList(dgt_filepath))
zone_id_key = {'D': 'Differentiation Zone', 'E': 'Elongation Zone',\
'M': 'Meristem'}
zone = zone_id_key[WT_zone_id]
#Generate venn diagram
venn = venn2_unweighted([WT_expr, dgt_expr], set_labels=("WT", "dgt"))
plt.title(str(zone))
plt.savefig(os.path.join(venn_output_location, venn_output_filename))
#Create gene lists for each section of venn diagram
intersections_dict = {'WT_dgt': WT_expr.intersection(dgt_expr),\
'WT_only': WT_expr.difference(dgt_expr), 'dgt_only': dgt_expr.difference(WT_expr)}
for inter_key in intersections_dict:
filename = inter_key + '_' + WT_zone_id + '_' + WT_cutoff_id + '.txt'
filepath = os.path.join(genelist_output_location, filename)
IF = open(filepath, 'w')
for member in intersections_dict[inter_key]:
IF.write(member + '\n')
IF.close()
help='Each one representing one set. [2 or 3 files].')
parser.add_argument("-o",
"--output",
required=True,
help='Output basename to write the figure.')
args = parser.parse_args()
if len(args.inputfiles) < 2 or len(args.inputfiles) > 3:
print("Please provide at least 2 sets to compare.")
exit(1)
list_sets = []
labels = []
for file in args.inputfiles:
s = set()
l = os.path.basename(file).split(".")[0]
labels.append(l)
with open(file, 'r') as infile:
for line in infile:
s.add(line.rstrip())
infile.close()
list_sets.append(s)
if len(list_sets) == 2:
plt.figure(figsize=(7, 4))
venn2_unweighted(list_sets, labels)
elif len(list_sets) == 3:
plt.figure(figsize=(9, 5))
venn3_unweighted(list_sets, labels)
plt.savefig(args.output + ".png")
sub=[] for condition_nb in set_to_title.values(): if condition_nb in venn_diagram: sub.append(venn_diagram[condition_nb]) else: sub.append(0) # print tuple(sub) # avec les 3 fichiers test, resultat attendu #Only in test1 Only in test2 Common test1 test2 Only in test3 Common test1 test3 Common test2 test3 Common test1 test2 test3 #(8, 6, 1, 7, 0, 2, 0) plt.figure(figsize=(14,10)) # first number : width , second number : height if len(liste_sets) == 2: v = venn2_unweighted(subsets = tuple(sub), set_labels = (liste_sets[0], liste_sets[1])) elif len(liste_sets) == 3: v = venn3_unweighted(subsets = tuple(sub), set_labels = (liste_sets[0], liste_sets[1], liste_sets[2])) for text in v.set_labels: # file name size text.set_fontsize(12) for text in v.subset_labels: # numbers inside circles size text.set_fontsize(16) if args.venn_title is not None : plt.title(args.venn_title) # display title #liste_sets_string = ', '.join(liste_sets) #print liste_sets_string #plt.title(liste_sets_string)
deetcont = pd.read_csv("Genes/deetcont.csv", header=None)
perm = pd.read_csv("Genes/perm.csv", header=None)
permcont = pd.read_csv("Genes/permcont.csv", header=None)
genes = pd.read_csv("Genes/listofgenes.csv", header=None)
def compare(s1, s2):
c = len(list(set(s1.iloc[:, 0]) & set(s2.iloc[:, 0])))
return (c)
diff = compare(deet, genes)
len(deet)
fig, axes = plt.subplots(2)
v1 = venn2_unweighted(subsets=(len(deet), len(deetcont),
len(genes) - len(deetcont) - len(deet)),
set_labels=("Up Regulated", "Down Regulated"),
ax=axes[0])
v2 = venn2_unweighted(subsets=(len(perm), len(permcont),
len(genes) - len(permcont) - len(perm)),
set_labels=("Up Regulated", "Down Regulated"),
ax=axes[1])
axes[0].set_title("Deet")
axes[1].set_title("Perm")
fig.tight_layout()
fig.savefig("VennDiagrams/RevisedDiagrams/DeetnPerm.png",
bbox_inxhes="tight",
dpi=250)
plt.show()
#%% fig3
def feature_overlap(self,
exclude=None,
weighted=True,
label_fontsize=10,
count_fontsize=10,
save_path=None):
"""
Plots a venn diagram illustrating the overlap in features between the datasets.
:param save_path: Path to which venn diagram shall be saved.
:return:
"""
feat_set = self.get_feature_sets(exclude)
# Plotting when two datasets are compared
if len(self) == 2:
# set variables needed to assign new color scheme
colors = ["blue", "green"]
ids = ["A", "B"]
plt.figure(figsize=[3, 4])
if weighted:
# create circles
v = mv.venn2(feat_set, set_labels=self.df_names)
# create lines around circles
circles = mv.venn2_circles(feat_set)
else:
# create circles
v = mv.venn2_unweighted(feat_set, set_labels=self.df_names)
# create lines around circles
circles = mv.venn2_circles(subsets=(1, 1, 1))
# Plotting when three datasets are compared
elif len(self) == 3:
# set variables needed to assign new color scheme
colors = ["blue", "green", "purple"]
ids = ["A", "B", "001"]
# create cirlces
v = mv.venn3_unweighted(feat_set, set_labels=self.df_names)
# create lines around circles
circles = mv.venn3_circles(subsets=(1, 1, 1, 1, 1, 1, 1))
else:
raise ValueError(
"Too many datasets in DataCollection. Venn diagram only supported for up to 3 datasets."
)
# set colors for the circles in venn diagram
for df_name, color in zip(ids, colors):
v.get_patch_by_id(df_name).set_color(color)
# reduce line width around circles
for c in circles:
c.set_lw(1.0)
# change font size of dataset labels
for text in v.set_labels:
text.set_fontsize(label_fontsize)
# change font size of displayed counts
for text in v.subset_labels:
text.set_fontsize(count_fontsize)
plt.title("Feature Overlap")
if save_path:
plt.savefig(save_path, dpi=300)
else:
plt.show()
figure_pida_depErangE, axes_pida_depErangE = plt.subplots(1, 2)
subsets_pida_depErangE_p = (nooverlap_pida_depErangeE_p0,
nooverlap_pida_depErangeE_p1,
overlap_pida_depErangeE_p01)
areas_pida_depErangE_p = [1, 1, 1]
for i in range(0, 3):
if subsets_pida_depErangE_p[i] < 0.03 * true_protons:
areas_pida_depErangE_p[i] = 0.03 * true_protons
else:
areas_pida_depErangE_p[i] = subsets_pida_depErangE_p[i]
venn2_unweighted(subsets_pida_depErangE_p,
set_labels=('PIDA', r'$E_{DEP} - E_{RANGE}$'),
ax=axes_pida_depErangE[0],
subset_areas=areas_pida_depErangE_p)
subsets_pida_depErangE_mu = (nooverlap_pida_depErangeE_mu0,
nooverlap_pida_depErangeE_mu1,
overlap_pida_depErangeE_mu01)
areas_pida_depErangE_mu = [1, 1, 1]
for i in range(0, 3):
if subsets_pida_depErangE_mu[i] < 0.03 * true_muons:
areas_pida_depErangE_mu[i] = 0.03 * true_muons
else:
areas_pida_depErangE_mu[i] = subsets_pida_depErangE_mu[i]
venn2_unweighted(subsets_pida_depErangE_mu,
set_labels=('PIDA', r'$E_{DEP} - E_{RANGE}$'),
# creating dataframe from dictionary
dataframe = pd.DataFrame.from_dict(dic_df).drop_duplicates(subset='name')
pd.set_option('display.width', None)
print(dataframe)
valence_vals = dataframe['valence'].tolist()
less_count, more_count, middle_count = 0, 0, 0
for num in valence_vals:
if num >= 0 and num < 0.5:
less_count += 1
elif num >= 0.5 and num < 0.6:
middle_count += 1
else:
more_count += 1
venn2_unweighted(subsets=(less_count, more_count, middle_count),
set_labels=('Low Spirit', 'High Spirit'),
set_colors=('navy', 'lime'),
alpha=0.5)
dataframe.plot.line(x='time', y=['danceability', 'energy', 'valence'])
plt.xticks(rotation=90)
plt.show()
# if the graph is erratic, thay maybe because of streaming of a particular artist/ album,
# since an album contains a mixture of sad and energetic songs. so we take a look
# at the number of unique artists and albums in the history :
print(dataframe['album'].value_counts(ascending=False))
print(dataframe['artist'].value_counts(ascending=False))