van Hove plotting problem

[Resonanz]

I am trying to get the van Hove function to work.

In the walkthrough I have added a little code to call the van Hove function as follows:

#existing in walkthrough
t2 = t1[((t1['mass'] > 300) & (t1['size'] < 3.5) &
         (t1['ecc'] < 0.1))]

#added to walkthrough
pos = t2.set_index(['frame','particle'])['x'].unstack()
vh = tp.vanhove(pos, lagtime=2)

I get this error for a span of lagtimes:

ValueError: ('range parameter must be finite.', 'occurred at index 0.0')

which is coming from this line in the van Hove function:

vh = disp.apply(lambda x: Series(np.histogram(x, bins=global_bins, density=True)[0]))

Any help? Thanks !

[Resonanz]

As a follow-on, I am finding the code really hard to follow (well, pandas is quite complicated). For example. code like this does not seem to be explained by the histogram documentation:

global_bins = np.histogram(values, bins=bins)[1]

So, what exactly does the [1] at the end mean? Is this an axis? Is the [1] old code that needs replacing with axis=1?

Thanks again.

[Resonanz]

This is the error code by the way:

ValueError                                Traceback (most recent call last)
<ipython-input-47-fd307fb7b8ec> in <module>()
     29 #print(x)
     30 #print('global_bins = ' + str(global_bins))
---> 31 vh = disp.apply(lambda x: Series(np.histogram(x, bins=global_bins, density=True)[0]))
     32 #vh = Series(np.histogram(x, bins=global_bins, density=True)[0])
     33 #vh.index = global_bins[:-1]

C:\Users\rlward\AppData\Local\Continuum\Anaconda3\lib\site-packages\pandas\core\frame.py in apply(self, func, axis, broadcast, raw, reduce, args, **kwds)
   4059                     if reduce is None:
   4060                         reduce = True
-> 4061                     return self._apply_standard(f, axis, reduce=reduce)
   4062             else:
   4063                 return self._apply_broadcast(f, axis)

C:\Users\rlward\AppData\Local\Continuum\Anaconda3\lib\site-packages\pandas\core\frame.py in _apply_standard(self, func, axis, ignore_failures, reduce)
   4155             try:
   4156                 for i, v in enumerate(series_gen):
-> 4157                     results[i] = func(v)
   4158                     keys.append(v.name)
   4159             except Exception as e:

<ipython-input-47-fd307fb7b8ec> in <lambda>(x)
     29 #print(x)
     30 #print('global_bins = ' + str(global_bins))
---> 31 vh = disp.apply(lambda x: Series(np.histogram(x, bins=global_bins, density=True)[0]))
     32 #vh = Series(np.histogram(x, bins=global_bins, density=True)[0])
     33 #vh.index = global_bins[:-1]

C:\Users\rlward\AppData\Local\Continuum\Anaconda3\lib\site-packages\numpy\lib\function_base.py in histogram(a, bins, range, normed, weights, density)
    503     if not np.all(np.isfinite([mn, mx])):
    504         raise ValueError(
--> 505             'range parameter must be finite.')
    506     if mn == mx:
    507         mn -= 0.5

ValueError: ('range parameter must be finite.', 'occurred at index 0.0')

[tacaswell]

np.histogram returns 2 values in a tuple (https://docs.scipy.org/doc/numpy/reference/generated/numpy.histogram.html). The [1] is just selecting the second value.

Those exceptions look like some invalid data (np.nan or np.inf) is getting into the histogram call.

[Resonanz]

There are NaN vales in the data but I think these are processed out. Would it be possible to see an example of some working code, or should my code above work? If I add those lines into the Walkthrough, should they work?

[caspervdw]

This needs some documentation.

[pipoUrdaneta]

Hi all

I signed up to this website just a while ago feeling pretty desperate about exactly the same problem Resonanz posted. Has any of you gotten around this problem yet?

PS: I'm totally new to coding (started learning Python from scratch just over a month ago).

[Resonanz]

I have just started a new microrheology project so I'm keen (again) to get this working.

[pipoUrdaneta]

My seniors advised me to uninstall Anaconda3-5.x, and install Anaconda3-4.0.0. Did it just this morning and now the van Hove command worked.

[caspervdw]

@pipoUrdaneta Good that it works for you! I wonder what your pandas and numpy version were with the respective installations. These can be shown easily using

import trackpy as tp
print(tp.diag.dependencies())

@danielballan Would be the one to provide a working example as he wrote the function.

For sure we need a test to at least run the vanhove function with different library versions.

[Resonanz]

If you have code that now works, could you please post a working example that others can try? Thanks.

[pipoUrdaneta]

@caspervdw Thanks!
import trackpy as tp
print(tp.diag.dependencies())
gives this:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-1-1f4075b94f34> in <module>()
      1 import trackpy as tp
----> 2 print(tp.diag.dependencies())

AttributeError: module 'trackpy' has no attribute 'diag'

But for pandas and numpy in my current Anaconda3-4.0.0 installation:

import numpy as np
np.__version__

gives
'1.10.4'

and

import pandas as pd
pd.__version__ 

gives
'0.18.0'

The Anaconda version previously used was 5.2.0, with Python 3.6.
I looked up the packages from that installation here:
https://docs.anaconda.com/anaconda/packages/py3.6_win-32
and for Numpy it's 1.14.3,
while for Pandas it's 0.23.0

[pipoUrdaneta]

@Resonanz will do, although for the above I did just follow the van Hove command in TrackPy's documentation,
http://soft-matter.github.io/trackpy/dev/generated/trackpy.motion.vanhove.html

pos = traj.set_index(['frame', 'particle'])['x'].unstack() # particles as columns
vh = tp.vanhove(pos)

(and not forgetting to specify the lagtime argument after pos)

[danielballan]

@danielballan Would be the one to provide a working example as he wrote the function.

Fair. See #516.