make benchmark easier to reproduce

.gitignore

@@ -2,4 +2,4 @@
 *~
 *.png
 *.json
-cache/
+

README.rst

@@ -32,7 +32,7 @@ MatFile    Matlab .mat file         ☒ ☒ ☒    ☒ ☒ ☐    ☒ ☒ ☐
 * + Rating refers to using a semi-popular package (probably scipy), as opposed to only python and numpy.
 * ++ Very easy (☒☒☒) with an unpopular and/or dedicated package, but the rating refers to only python and numpy.
 
-csv and NPY were done with `numpy`_; json was done with `pyjson_tricks`_; png was done with `imgarray`_; fortran unformatted was done with `fortranfile`_; matlab was done with `scipy`_; pickle, base64 and gzipping were done with python built-ins.
+You can install all dependencies using `pip install -r rements.pip`. csv and NPY were done with `numpy`_; json was done with `pyjson_tricks`_; png was done with `imgarray`_; fortran unformatted was done with `fortranfile`_; matlab was done with `scipy`_; pickle, base64 and gzipping were done with python built-ins. `pandas`_ is needed for stata (not ready yet) statai files. Seaborn is needed for plotting. You can install all dependencies using `pip install requirements.pip`
 
 Results
 ---------------------------------------
@@ -54,7 +54,7 @@ Sparse random matrix
 Real data
 =======================================
 
-Scattering probabilities for hydrogen and carbon monoxide (many doubles between 0 and 1, most close to 0).
+Scattering probabilities for hydrogen and carbon monoxide (many doubles between 0 and 1, most close to 0). You can easily overwrite this by your own file in `testdata.csv`.
 
 .. image:: https://raw.githubusercontent.com/mverleg/array_storage_benchmark/master/result/bm_example.png
 
@@ -67,7 +67,7 @@ Pull requests with other methods (serious or otherwise) are welcome! There might
 .. _`numpy`: https://docs.scipy.org/doc/numpy/reference/generated/numpy.loadtxt.html
 .. _`pyjson_tricks`: https://github.com/mverleg/pyjson_tricks
 .. _`imgarray`: https://github.com/mverleg/imgarray
-.. _`fortranfile_`: https://pypi.python.org/pypi/fortranfile/0.2.1
+.. _`fortranfile`: https://pypi.python.org/pypi/fortranfile/0.2.1
 .. _`scipy`: https://docs.scipy.org/doc/scipy-0.18.1/reference/generated/scipy.io.loadmat.html#scipy.io.loadmat
 
 

benchmark.py

@@ -38,11 +38,6 @@ def __init__(self, cls, data, data_name=None, reps=50):
 	@property
 	def save_time(self):
 		assert self.done
-		# mn, mx, st = mean(tuple(inst.save_time for inst in self.done)), max(tuple(inst.save_time for inst in self.done)), std(tuple(inst.save_time for inst in self.done))
-		# if mx > 2 * mn:  #todo
-		# 	print('max', self.cls.__name__, mn, mx)
-		# if st > 0.2 * mn:
-		# 	print('std', self.cls.__name__, mn, st)
 		return mean(tuple(inst.save_time for inst in self.done))
 
 	@property
@@ -94,14 +89,13 @@ def random_data(size, is_sparse=False, is_big=True):
 	else:
 		arr = rs.rand(*size).astype('float64')
 	if is_big:
-		# arr = (2 * arr - 1) * 1.7976931348623157e+308
 		# don't use the full range, since some formats (Stata) uses the highest values for special meanings.
 		arr = (arr - 0.5) * 1.7976931348623157e+308
 	return arr
 
 
 def load_example_data():
-	return loadtxt(join(environ['HOME'], 'testdata.csv'), delimiter=',')
+	return loadtxt('testdata.csv', delimiter=',')
 
 
 if __name__ == '__main__':

requirements.pip

@@ -0,0 +1,9 @@
+numpy
+scipy
+json-tricks
+imgarray
+fortranfile
+matplotlib
+seaborn
+pandas
+