def describe_image(image_path):
try:
description = describe(image_path)
translation = translate(description[0])
return (translation, description[1])
except:
return "Não consegui decifrar"
def cmd(file, pop_variance: float, confidence: float):
with open(file) as f:
reader = csv.reader(f)
# 2-dimensional array, but each row can be different number of elements.
contents = [[float(v) for v in row] for row in reader]
for i, row in enumerate(contents):
n, sample_mean, sample_var, _ = describe(row)
result = estimate_all(n, sample_mean, sample_var, pop_variance,
confidence)
print("系列{}".format(i + 1))
print(result)
def main():
parser = argparse.ArgumentParser(description='j\'suis un choixpeau magic')
parser.add_argument(
'-s',
'--scrypt',
type=str,
dest="scrypt",
choices=['describe', 'histogram', 'scatter_plot', 'pair_plot'],
help="function scrypt")
parser.add_argument(type=str, dest="dataset", help="describe dataset")
opt = parser.parse_args()
if (opt.scrypt == 'describe'):
descr.describe(opt.dataset)
elif (opt.scrypt == 'histogram'):
histo.histogram(opt.dataset)
elif (opt.scrypt == 'scatter_plot'):
scatter.scatter(opt.dataset)
elif (opt.scrypt == 'pair_plot'):
pair.pair_plot(opt.dataset)
def caller(path):
start_time = time.time()
df = pd.read_csv(path)
bdf = bpd.read_csv(path)
parent_dict = describe(df, bdf)
end_time = time.time()
seconds = end_time - start_time
return parent_dict, seconds
def _valueString(value, verbose=0):
"""Returns name and, for some types, value of the variable as a string."""
t = type(value)
vstr = t.__name__
if issubclass(t, str):
if len(value) > 42:
vstr = vstr + ", value = " + ` value[:39] ` + '...'
else:
vstr = vstr + ", value = " + ` value `
elif issubclass(t, _listTypes):
return "%s [%d entries]" % (vstr, len(value))
elif (PY3K and issubclass(t, io.IOBase)) or \
(not PY3K and issubclass(t, file)):
vstr = vstr + ", " + ` value `
elif issubclass(t, _numericTypes):
vstr = vstr + ", value = " + ` value `
elif _isinstancetype(value):
cls = value.__class__
if cls.__module__ == '__main__':
vstr = 'instance of class ' + cls.__name__
else:
vstr = 'instance of class ' + cls.__module__ + '.' + cls.__name__
elif issubclass(t, _functionTypes + _methodTypes):
# try using Fredrik Lundh's describe on functions
try:
vstr = vstr + ' ' + describe.describe(value)
try:
if verbose and value.__doc__:
vstr = vstr + "\n" + value.__doc__
except AttributeError:
pass
except (AttributeError, TypeError):
# oh well, just have to live with type string alone
pass
elif issubclass(t, _numpyArrayType):
vstr = vstr + " " + str(value.dtype) + "["
for k in range(len(value.shape)):
if k:
vstr = vstr + "," + ` value.shape[k] `
else:
vstr = vstr + ` value.shape[k] `
vstr = vstr + "]"
else:
# default -- just return the type
pass
return vstr
def test_estimation(self):
data = [100, 70, 30, 60, 50]
length, mean, variance, _ = describe.describe(data)
pop_variance = 625
confidence = 0.95
bottom, top = estimation.interval_estimate_mean_with_pop_variance(
mean, pop_variance, length, confidence)
self.assertAlmostEqual(bottom, 40.1, places=1)
self.assertAlmostEqual(top, 83.9, places=1)
bottom, top = estimation.interval_estimate_mean_without_pop_variance(
mean, variance, length, confidence)
self.assertAlmostEqual(bottom, 29.9, places=1)
self.assertAlmostEqual(top, 94.1, places=1)
bottom, top = estimation.interval_estimate_variance(
variance, length, confidence)
self.assertAlmostEqual(bottom, 241, places=0)
self.assertAlmostEqual(top, 5532, places=0)
family="Courier New, monospace",
size=8,
color="#7f7f7f"
)
)
fig.update_xaxes(title_text="Scores", row=row, col=column)
fig.update_yaxes(title_text="Frequency", row=row, col=column)
# Overlay both histograms
fig.update_layout(barmode='overlay')
# Reduce opacity to see both histograms
fig.update_traces(opacity=0.75)
if __name__ == "__main__":
df = readData(sys.argv)
df_numerical = describe(df)
titles = tuple(df_numerical.columns)
fig = make_subplots(
rows=ROWS, cols=COLUMNS,
subplot_titles=titles,
specs=[
# row 1
[{"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True},{"secondary_y": True} ],
# row 2
[{"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True},{"secondary_y": True} ],
# row 3
[{"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True}, {"secondary_y": True},{"secondary_y": True} ]
])
r = 1
index = 0
_len = len(titles)
import sys
houses = {"Ravenclaw": 1, "Slytherin": 2, "Gryffindor": 3, "Hufflepuff": 4}
def plot_hist(data, col):
plt.figure()
plt.title(data.columns[col])
data = data.to_numpy()
for i in range(1, 5):
curr_house = []
for row in data:
if row[1] == i and not np.isnan(row[col]):
curr_house.append(row[col])
plt.hist(curr_house, alpha=0.5)
plt.show()
if __name__ == "__main__":
np.set_printoptions(suppress=True)
try:
data = pd.read_csv("resources/dataset_train.csv")
except:
sys.exit("Error")
data["Hogwarts House"].replace(houses, inplace=True)
data = data.select_dtypes('number')
metrics = describe.describe(data.to_numpy())
metrics = metrics.tolist()
col = metrics[5].index(min(metrics[5]))
plot_hist(data, col)
def describe(self):
describe.describe(self.df)
import torch
import numpy as np
from describe import describe
# Torch randoms
describe(torch.rand(2,3)) # uniform random
describe(torch.randn(2,3)) # random normal
# Torch creating tensors
describe(torch.zeros(2,3))
x = torch.ones(2,3)
describe(x)
x.fill_(5) # All _ methods refer to in-place operations
describe(x)
x = torch.Tensor([[1,2,3],
[4,5,6]])
describe(x)
npy = np.random.rand(2,3)
describe(torch.from_numpy(npy))
# Torch tensor operations
describe(torch.add(x,x))
describe(x + x)
x = torch.arange(6)
describe(x)
x = x.view(2,3)
describe(x)
