def f():
if method == "less than":
result = data.loc[data[column] < value, ]
elif method == "less than or equal to":
result = data.loc[data[column] <= value, ]
elif method == "equal to":
result = data.loc[data[column] == value, ]
elif method == "not equal to":
result = data.loc[data[column] != value, ]
elif method == "greater than or equal to":
result = data.loc[data[column] >= value, ]
elif method == "greater than":
result = data.loc[data[column] > value, ]
print("=" * 80)
print(
f"Running Rowfilter: Keep rows where {repr(column)} is {method} {repr(value)}"
)
print("-" * 80)
print(
f"Kept {len(result):,} of {len(data):,} rows ({len(result)/len(data):.2%})"
)
print("=" * 80)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = clarite.modify.colfilter_percent_zero(
data, filter_percent, skip, only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = data.copy(
deep=True
) # This function works in-place, so a copy must be created first
clarite.analyze.add_corrected_pvalues(result)
if data_name is None:
return result
else:
return Dataset(data_name, "ewas_result", result)
def f():
result = clarite.modify.transform(data=data,
transform_method=method,
skip=skip,
only=only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = clarite.modify.recode_values(
data=data,
replacement_dict=replacement_dict,
skip=skip,
only=only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = clarite.modify.remove_outliers(data=data,
method=method,
cutoff=cutoff,
skip=skip,
only=only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = clarite.describe.get_types(data).reset_index()
result.columns = ["variable", "type"]
return Dataset(data_name, "datatypes", result)
def f():
result = clarite.describe.freq_table(data)
return Dataset(data_name, "freqtable", result)
def f():
result = clarite.describe.correlations(data, threshold)
return Dataset(data_name, "correlations", result)
def f():
result = clarite.analyze.ewas(**kwargs)
return Dataset(data_name, "ewas_result", result)
def f():
result = clarite.modify.colfilter_min_n(data, n, skip, only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
result = clarite.modify.merge_observations(top, bottom)
return Dataset(data_name, "dataset", result)
def f():
result = clarite.modify.merge_variables(left, right, how)
return Dataset(data_name, "dataset", result)
def f():
result = clarite.describe.percent_na(data)
return Dataset(data_name, "percentna", result)
def f():
result = clarite.describe.skewness(data, dropna)
return Dataset(data_name, "skewness", result)
def f():
result = clarite.modify.rowfilter_incomplete_obs(data, skip, only)
if data_name is None:
return result
else:
return Dataset(data_name, "dataset", result)
def f():
if kind == "CSV":
df = clarite.load.from_csv(filename, index_col)
elif kind == "TSV":
df = clarite.load.from_tsv(filename, index_col)
return Dataset(data_name, "dataset", df)
