def test_pandas_series_loading(self):
"""Pandas Series objects are correctly loaded"""
# Test valid series types
name = ['_x', ' name']
length = [0, 1, 2]
index_key = [None, 'ix', 1]
index_types = ['int', 'char', 'datetime', 'Timestamp']
value_key = [None, 'x', 1]
value_types = ['int', 'char', 'datetime', 'Timestamp', 'float',
'numpy float', 'numpy int']
series_info = product(name, length, index_key, index_types,
value_key, value_types)
for n, l, ikey, itype, vkey, vtype in series_info:
index = sequences[itype](l)
series = pd.Series(sequences[vtype](l), index=index, name=n,)
vkey = vkey or series.name
expected = [{'idx': Data.serialize(i), 'col': vkey,
'val': Data.serialize(v)}
for i, v in zip(index, series)]
data = Data.from_pandas(series, name=n, series_key=vkey)
nt.assert_list_equal(expected, data.values)
nt.assert_equal(n, data.name)
data.to_json()
# Missing a name
series = pd.Series(np.random.randn(10))
data = Data.from_pandas(series)
nt.assert_equal(data.name, 'table')
def test_pandas_series_loading(self):
"""Pandas Series objects are correctly loaded"""
# Test valid series types
name = ['_x', ' name']
length = [0, 1, 2]
index_key = [None, 'ix', 1]
index_types = ['int', 'char', 'datetime', 'Timestamp']
value_key = [None, 'x', 1]
value_types = ['int', 'char', 'datetime', 'Timestamp', 'float',
'numpy float', 'numpy int']
series_info = product(name, length, index_key, index_types,
value_key, value_types)
for n, l, ikey, itype, vkey, vtype in series_info:
index = sequences[itype](l)
series = pd.Series(sequences[vtype](l), index=index, name=n,)
vkey = series.name or vkey
expected = [{'idx': Data.serialize(i), 'col': vkey,
'val': Data.serialize(v)}
for i, v in zip(index, series)]
data = Data.from_pandas(series, name=n, series_key=vkey)
nt.assert_list_equal(expected, data.values)
nt.assert_equal(n, data.name)
data.to_json()
# Missing a name
series = pd.Series(np.random.randn(10))
data = Data.from_pandas(series)
nt.assert_equal(data.name, 'table')
def test_pandas_dataframe_loading(self):
# Simple columns/key_on tests
df = pd.DataFrame({'one': [1, 2, 3], 'two': [6, 7, 8],
'three': [11, 12, 13], 'four': [17, 18, 19]})
get_all = [{'col': 'four', 'idx': 0, 'val': 17},
{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'three', 'idx': 0, 'val': 11},
{'col': 'two', 'idx': 0, 'val': 6},
{'col': 'four', 'idx': 1, 'val': 18},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'three', 'idx': 1, 'val': 12},
{'col': 'two', 'idx': 1, 'val': 7},
{'col': 'four', 'idx': 2, 'val': 19},
{'col': 'one', 'idx': 2, 'val': 3},
{'col': 'three', 'idx': 2, 'val': 13},
{'col': 'two', 'idx': 2, 'val': 8}]
get1 = [{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'one', 'idx': 2, 'val': 3}]
get2 = [{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'two', 'idx': 0, 'val': 6},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'two', 'idx': 1, 'val': 7},
{'col': 'one', 'idx': 2, 'val': 3},
{'col': 'two', 'idx': 2, 'val': 8}]
getkey2 = [{'col': 'one', 'idx': 6, 'val': 1},
{'col': 'one', 'idx': 7, 'val': 2},
{'col': 'one', 'idx': 8, 'val': 3}]
getkey3 = [{'col': 'one', 'idx': 11, 'val': 1},
{'col': 'two', 'idx': 11, 'val': 6},
{'col': 'one', 'idx': 12, 'val': 2},
{'col': 'two', 'idx': 12, 'val': 7},
{'col': 'one', 'idx': 13, 'val': 3},
{'col': 'two', 'idx': 13, 'val': 8}]
val_all = Data.from_pandas(df)
val1 = Data.from_pandas(df, columns=['one'])
val2 = Data.from_pandas(df, columns=['one', 'two'])
key2 = Data.from_pandas(df, columns=['one'], key_on='two')
key3 = Data.from_pandas(df, columns=['one', 'two'], key_on='three')
nt.assert_list_equal(val_all.values, get_all)
nt.assert_list_equal(val1.values, get1)
nt.assert_list_equal(val2.values, get2)
nt.assert_list_equal(key2.values, getkey2)
nt.assert_list_equal(key3.values, getkey3)
# Missing a name
dataframe = pd.DataFrame(np.random.randn(10, 3))
data = Data.from_pandas(dataframe)
nt.assert_equal(data.name, 'table')
# Bad obj
nt.assert_raises(ValueError, Data.from_pandas, {})
def test_pandas_dataframe_loading(self):
"""Pandas DataFrame objects are correctly loaded"""
# name = ['_x']
# length = [0, 1, 2]
# index_key = [None, 'ix', 1]
# index_types = ['int', 'char', 'datetime', 'Timestamp']
# column_types = ['int', 'char', 'datetime', 'Timestamp']
# # Leaving out some basic types here because we're not worried about
# # serialization.
# value_types = [
# 'char', 'datetime', 'Timestamp', 'numpy float', 'numpy int']
# dataframe_info = product(
# name, length, length, index_key, index_types, column_types,
# value_types)
# for n, rows, cols, ikey, itype, ctype, vtype in dataframe_info:
# index = sequences[itype](rows)
# columns = sequences[ctype](cols)
# series = {
# c: pd.Series(sequences[vtype](rows), index=index, name=n)
# for c in columns}
# dataframe = pd.DataFrame(series)
# ikey = ikey or Data._default_index_key
# if cols == 0:
# expected = []
# else:
# expected = [
# dict([(ikey, Data.serialize(index[i]))] +
# [(str(c), Data.serialize(series[c][i]))
# for c in columns])
# for i in xrange(rows)]
# data = Data.from_pandas(dataframe, name=n, index_key=ikey)
# nt.assert_list_equal(expected, data.values)
# nt.assert_equal(n, data.name)
# data.to_json()
#Simple columns/key_on tests
df = pd.DataFrame({'one': [1,2,3], 'two': [6,7,8],
'three': [11, 12, 13], 'four': [17, 18, 19]})
get_all = [{'col': 'four', 'idx': 0, 'val': 17},
{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'three', 'idx': 0, 'val': 11},
{'col': 'two', 'idx': 0, 'val': 6},
{'col': 'four', 'idx': 1, 'val': 18},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'three', 'idx': 1, 'val': 12},
{'col': 'two', 'idx': 1, 'val': 7},
{'col': 'four', 'idx': 2, 'val': 19},
{'col': 'one', 'idx': 2, 'val': 3},
{'col': 'three', 'idx': 2, 'val': 13},
{'col': 'two', 'idx': 2, 'val': 8}]
get1 = [{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'one', 'idx': 2, 'val': 3}]
get2 = [{'col': 'one', 'idx': 0, 'val': 1},
{'col': 'two', 'idx': 0, 'val': 6},
{'col': 'one', 'idx': 1, 'val': 2},
{'col': 'two', 'idx': 1, 'val': 7},
{'col': 'one', 'idx': 2, 'val': 3},
{'col': 'two', 'idx': 2, 'val': 8}]
getkey2 = [{'col': 'one', 'idx': 6, 'val': 1},
{'col': 'one', 'idx': 7, 'val': 2},
{'col': 'one', 'idx': 8, 'val': 3}]
getkey3 = [{'col': 'one', 'idx': 11, 'val': 1},
{'col': 'two', 'idx': 11, 'val': 6},
{'col': 'one', 'idx': 12, 'val': 2},
{'col': 'two', 'idx': 12, 'val': 7},
{'col': 'one', 'idx': 13, 'val': 3},
{'col': 'two', 'idx': 13, 'val': 8}]
val_all = Data.from_pandas(df)
val1 = Data.from_pandas(df, columns=['one'])
val2 = Data.from_pandas(df, columns=['one', 'two'])
key2 = Data.from_pandas(df, columns=['one'], key_on='two')
key3 = Data.from_pandas(df, columns=['one', 'two'], key_on='three')
nt.assert_list_equal(val_all.values, get_all)
nt.assert_list_equal(val1.values, get1)
nt.assert_list_equal(val2.values, get2)
nt.assert_list_equal(key2.values, getkey2)
nt.assert_list_equal(key3.values, getkey3)
# Missing a name
dataframe = pd.DataFrame(np.random.randn(10, 3))
data = Data.from_pandas(dataframe)
nt.assert_equal(data.name, 'table')
#Bad obj
nt.assert_raises(ValueError, Data.from_pandas, {})
def test_pandas_dataframe_loading(self):
"""Pandas DataFrame objects are correctly loaded"""
# name = ['_x']
# length = [0, 1, 2]
# index_key = [None, 'ix', 1]
# index_types = ['int', 'char', 'datetime', 'Timestamp']
# column_types = ['int', 'char', 'datetime', 'Timestamp']
# # Leaving out some basic types here because we're not worried about
# # serialization.
# value_types = [
# 'char', 'datetime', 'Timestamp', 'numpy float', 'numpy int']
# dataframe_info = product(
# name, length, length, index_key, index_types, column_types,
# value_types)
# for n, rows, cols, ikey, itype, ctype, vtype in dataframe_info:
# index = sequences[itype](rows)
# columns = sequences[ctype](cols)
# series = {
# c: pd.Series(sequences[vtype](rows), index=index, name=n)
# for c in columns}
# dataframe = pd.DataFrame(series)
# ikey = ikey or Data._default_index_key
# if cols == 0:
# expected = []
# else:
# expected = [
# dict([(ikey, Data.serialize(index[i]))] +
# [(str(c), Data.serialize(series[c][i]))
# for c in columns])
# for i in xrange(rows)]
# data = Data.from_pandas(dataframe, name=n, index_key=ikey)
# nt.assert_list_equal(expected, data.values)
# nt.assert_equal(n, data.name)
# data.to_json()
#Simple columns/key_on tests
df = pd.DataFrame({
'one': [1, 2, 3],
'two': [6, 7, 8],
'three': [11, 12, 13],
'four': [17, 18, 19]
})
get_all = [{
'col': 'four',
'idx': 0,
'val': 17
}, {
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'three',
'idx': 0,
'val': 11
}, {
'col': 'two',
'idx': 0,
'val': 6
}, {
'col': 'four',
'idx': 1,
'val': 18
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'three',
'idx': 1,
'val': 12
}, {
'col': 'two',
'idx': 1,
'val': 7
}, {
'col': 'four',
'idx': 2,
'val': 19
}, {
'col': 'one',
'idx': 2,
'val': 3
}, {
'col': 'three',
'idx': 2,
'val': 13
}, {
'col': 'two',
'idx': 2,
'val': 8
}]
get1 = [{
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'one',
'idx': 2,
'val': 3
}]
get2 = [{
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'two',
'idx': 0,
'val': 6
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'two',
'idx': 1,
'val': 7
}, {
'col': 'one',
'idx': 2,
'val': 3
}, {
'col': 'two',
'idx': 2,
'val': 8
}]
getkey2 = [{
'col': 'one',
'idx': 6,
'val': 1
}, {
'col': 'one',
'idx': 7,
'val': 2
}, {
'col': 'one',
'idx': 8,
'val': 3
}]
getkey3 = [{
'col': 'one',
'idx': 11,
'val': 1
}, {
'col': 'two',
'idx': 11,
'val': 6
}, {
'col': 'one',
'idx': 12,
'val': 2
}, {
'col': 'two',
'idx': 12,
'val': 7
}, {
'col': 'one',
'idx': 13,
'val': 3
}, {
'col': 'two',
'idx': 13,
'val': 8
}]
val_all = Data.from_pandas(df)
val1 = Data.from_pandas(df, columns=['one'])
val2 = Data.from_pandas(df, columns=['one', 'two'])
key2 = Data.from_pandas(df, columns=['one'], key_on='two')
key3 = Data.from_pandas(df, columns=['one', 'two'], key_on='three')
nt.assert_list_equal(val_all.values, get_all)
nt.assert_list_equal(val1.values, get1)
nt.assert_list_equal(val2.values, get2)
nt.assert_list_equal(key2.values, getkey2)
nt.assert_list_equal(key3.values, getkey3)
# Missing a name
dataframe = pd.DataFrame(np.random.randn(10, 3))
data = Data.from_pandas(dataframe)
nt.assert_equal(data.name, 'table')
#Bad obj
nt.assert_raises(ValueError, Data.from_pandas, {})
def test_pandas_dataframe_loading(self):
# Simple columns/key_on tests
df = pd.DataFrame({
'one': [1, 2, 3],
'two': [6, 7, 8],
'three': [11, 12, 13],
'four': [17, 18, 19]
})
get_all = [{
'col': 'four',
'idx': 0,
'val': 17
}, {
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'three',
'idx': 0,
'val': 11
}, {
'col': 'two',
'idx': 0,
'val': 6
}, {
'col': 'four',
'idx': 1,
'val': 18
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'three',
'idx': 1,
'val': 12
}, {
'col': 'two',
'idx': 1,
'val': 7
}, {
'col': 'four',
'idx': 2,
'val': 19
}, {
'col': 'one',
'idx': 2,
'val': 3
}, {
'col': 'three',
'idx': 2,
'val': 13
}, {
'col': 'two',
'idx': 2,
'val': 8
}]
get1 = [{
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'one',
'idx': 2,
'val': 3
}]
get2 = [{
'col': 'one',
'idx': 0,
'val': 1
}, {
'col': 'two',
'idx': 0,
'val': 6
}, {
'col': 'one',
'idx': 1,
'val': 2
}, {
'col': 'two',
'idx': 1,
'val': 7
}, {
'col': 'one',
'idx': 2,
'val': 3
}, {
'col': 'two',
'idx': 2,
'val': 8
}]
getkey2 = [{
'col': 'one',
'idx': 6,
'val': 1
}, {
'col': 'one',
'idx': 7,
'val': 2
}, {
'col': 'one',
'idx': 8,
'val': 3
}]
getkey3 = [{
'col': 'one',
'idx': 11,
'val': 1
}, {
'col': 'two',
'idx': 11,
'val': 6
}, {
'col': 'one',
'idx': 12,
'val': 2
}, {
'col': 'two',
'idx': 12,
'val': 7
}, {
'col': 'one',
'idx': 13,
'val': 3
}, {
'col': 'two',
'idx': 13,
'val': 8
}]
val_all = Data.from_pandas(df)
val1 = Data.from_pandas(df, columns=['one'])
val2 = Data.from_pandas(df, columns=['one', 'two'])
key2 = Data.from_pandas(df, columns=['one'], key_on='two')
key3 = Data.from_pandas(df, columns=['one', 'two'], key_on='three')
nt.assert_list_equal(val_all.values, get_all)
nt.assert_list_equal(val1.values, get1)
nt.assert_list_equal(val2.values, get2)
nt.assert_list_equal(key2.values, getkey2)
nt.assert_list_equal(key3.values, getkey3)
# Missing a name
dataframe = pd.DataFrame(np.random.randn(10, 3))
data = Data.from_pandas(dataframe)
nt.assert_equal(data.name, 'table')
# Bad obj
nt.assert_raises(ValueError, Data.from_pandas, {})
