# how to get all data from all monitors? **URL:** \ **Category:** Science, Projects, and Showcases\ **Created:** [18 February 2021 17:50 UTC](https://brian.discourse.group/t/how-to-get-all-data-from-all-monitors/302 "2021-02-18T17:50:08Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![Jitender](https://yyz2.discourse-cdn.com/free1/user_avatar/brian.discourse.group/jitender/32/151_2.png) [@Jitender](https://brian.discourse.group/u/Jitender)\ **Post date:** [18 February 2021 17:50 UTC](https://brian.discourse.group/t/how-to-get-all-data-from-all-monitors/302/1 "2021-02-18T17:50:08Z")
I don’t know much about pickle files. As you gave examples to get data from Monitors for only in Pickle file. But i want to get data in .csv, .txt file. I tried for that but I was unable to get all data, in that file, very less data comes with dots. Please suggest me How can i get data from monitors in csv file. ---
**Author:** ![mstimberg](https://yyz2.discourse-cdn.com/free1/user_avatar/brian.discourse.group/mstimberg/32/11_2.png) [@mstimberg](https://brian.discourse.group/u/mstimberg)\ **Post date:** [18 February 2021 18:09 UTC](https://brian.discourse.group/t/how-to-get-all-data-from-all-monitors/302/2 "2021-02-18T18:09:42Z")
Hi and welcome to the forum. If you are only interested in a single recorded variable, you will get it as a standard numpy array froma monitor (or as a `Quantity` if it has physical units, but this is a subclass of the standard numpy array, see the [documentation on units](https://brian2.readthedocs.io/en/stable/user/units.html) for more details). You can then use a function like [numpy.savetxt](https://numpy.org/doc/stable/reference/generated/numpy.savetxt.html) to store it to a text file. Monitors (and groups of neurons, etc.) also provide the `get_states` method to get the values of several variables at once. By default it will return a dictionary, but it can also return a [pandas DataFrame](https://pandas.pydata.org/pandas-docs/stable/user_guide/dsintro.html#dataframe) which is convenient for storing as csv. Note that the pandas format has some limitations, though: it does not support physical units and it can only represent one-dimensional data. The latter restriction means that you cannot use it for variables recorded by a `StateMonitor`, since each variable is stored as a 2-d array (values per timestep and per neuron). But you can use it for example with a `SpikeMonitor`. E.g. to store the spike times and indices to a csv file, something like the following should work: ```auto spike_mon = SpikeMonitor(...) # ... run simulation # Get spike times/indices as a data frame: data = spike_mon.get_states(['t', 'i'], units=False, format='pandas') # Write data frame to disk as a CSV file: data.to_csv('spikes.csv', index=False) ``` Hope that makes things clearer. ---
**Author:** ![Jitender](https://yyz2.discourse-cdn.com/free1/user_avatar/brian.discourse.group/jitender/32/151_2.png) [@Jitender](https://brian.discourse.group/u/Jitender)\ **Post date:** [19 February 2021 04:32 UTC](https://brian.discourse.group/t/how-to-get-all-data-from-all-monitors/302/3 "2021-02-19T04:32:53Z")
> [@mstimberg](#): > > `('spikes.csv', index=False)` Sorry Sir, It’s not working for me. Actually I took 100 cells with connection probability 0.1. So I want to want to track all variable (e.g, voltage , conductance etc. and synaptic variables) I used ‘get\_states’ function with your example given above. But its not working for me. I want to track all variable for ‘dt’ time step. Say I run program for 3 Seconds with dt = 0.1 so i want 30000 data sets for each variable so i can use statistics to analysis those. If it is not possible please tell me other option I have given much time to understand brian2, I can’t leave it.My work is pending due to this problem. I hope you understand my problem. ---
**Author:** ![mstimberg](https://yyz2.discourse-cdn.com/free1/user_avatar/brian.discourse.group/mstimberg/32/11_2.png) [@mstimberg](https://brian.discourse.group/u/mstimberg)\ **Post date:** [19 February 2021 17:50 UTC](https://brian.discourse.group/t/how-to-get-all-data-from-all-monitors/302/4 "2021-02-19T17:50:37Z")
I’m afraid we cannot help you much more with this, given that this seems to be more about general data processing in Python and not so much about Brian. As I said earlier, when you record a variable for each time step for several neurons with a `StateMonitor`, the `StateMonitor` will give you either a 2-D `numpy.ndarray` or a 2-D `Quantity` (which is derived from `numpy.ndarray`). If you want to store this data to a file, you can use for example numpy’s `savetxt` function: ```auto # .... some model state_mon = StateMonitor(neurons, 'v', record=True) # recording variable 'v' from all neurons run(...) np.savetxt('v.txt', state_mon.v_.T) ``` I use `v_` instead of `v` above to get the values without units, but using `v` would also work since numpy would simply discard the units. An alternative would be to use something like `state_mon.T/mV` to get the values in mV. Note that I transposed the matrix with `.T` to get the more common format of one row for each time step and one column for each neuron – otherwise it would be the opposite. If you want to store the values from several variables in a file, then you’ll have to manually put them all into a single 2d array (with functions like `np.concatenate`), or create a pandas DataFrame with one column per variable+neuron. You could also write a little function that does this somewhat automatically for you. Of course you don’t need to save anything to a csv file to do statistics on it, you can do this with numpy/scipy/pandas directly. See e.g. [https://scipy-lectures.org/](https://scipy-lectures.org/) for teaching material related to this. Just as a trivial example: ```auto print(np.mean(state_mon.v, axis=1)) ``` This would give you the average membrane potential of each neuron (i.e. averaged over time).