1 .TH g_analyze 1 "Fri 18 Jan 2013" "" "GROMACS suite, VERSION 4.5.6"
3 g_analyze - analyzes data sets
8 .BI "\-f" " graph.xvg "
9 .BI "\-ac" " autocorr.xvg "
10 .BI "\-msd" " msd.xvg "
11 .BI "\-cc" " coscont.xvg "
12 .BI "\-dist" " distr.xvg "
13 .BI "\-av" " average.xvg "
14 .BI "\-ee" " errest.xvg "
15 .BI "\-bal" " ballisitc.xvg "
16 .BI "\-g" " fitlog.log "
18 .BI "\-[no]version" ""
28 .BI "\-errbar" " enum "
29 .BI "\-[no]integrate" ""
30 .BI "\-aver_start" " real "
32 .BI "\-[no]regression" ""
35 .BI "\-fitstart" " real "
36 .BI "\-fitend" " real "
37 .BI "\-smooth" " real "
38 .BI "\-filter" " real "
42 .BI "\-acflen" " int "
43 .BI "\-[no]normalize" ""
45 .BI "\-fitfn" " enum "
46 .BI "\-ncskip" " int "
47 .BI "\-beginfit" " real "
48 .BI "\-endfit" " real "
50 \&\fB g_analyze\fR reads an ASCII file and analyzes data sets.
51 \&A line in the input file may start with a time
52 \&(see option \fB \-time\fR) and any number of \fI y\fR\-values may follow.
53 \&Multiple sets can also be
54 \&read when they are separated by & (option \fB \-n\fR);
55 \&in this case only one \fI y\fR\-value is read from each line.
56 \&All lines starting with and @ are skipped.
57 \&All analyses can also be done for the derivative of a set
58 \&(option \fB \-d\fR).
61 \&All options, except for \fB \-av\fR and \fB \-power\fR, assume that the
62 \&points are equidistant in time.
65 \&\fB g_analyze\fR always shows the average and standard deviation of each
66 \&set, as well as the relative deviation of the third
67 \&and fourth cumulant from those of a Gaussian distribution with the same
71 \&Option \fB \-ac\fR produces the autocorrelation function(s).
72 \&Be sure that the time interval between data points is
73 \&much shorter than the time scale of the autocorrelation.
76 \&Option \fB \-cc\fR plots the resemblance of set i with a cosine of
77 \&i/2 periods. The formula is:
78 2 (integral from 0 to T of y(t) cos(i pi t) dt)2 / integral from 0 to T of y2(t) dt
80 \&This is useful for principal components obtained from covariance
81 \&analysis, since the principal components of random diffusion are
85 \&Option \fB \-msd\fR produces the mean square displacement(s).
88 \&Option \fB \-dist\fR produces distribution plot(s).
91 \&Option \fB \-av\fR produces the average over the sets.
92 \&Error bars can be added with the option \fB \-errbar\fR.
93 \&The errorbars can represent the standard deviation, the error
94 \&(assuming the points are independent) or the interval containing
95 \&90% of the points, by discarding 5% of the points at the top and
99 \&Option \fB \-ee\fR produces error estimates using block averaging.
100 \&A set is divided in a number of blocks and averages are calculated for
101 \&each block. The error for the total average is calculated from
102 \&the variance between averages of the m blocks B_i as follows:
103 \&error2 = sum (B_i \- B)2 / (m*(m\-1)).
104 \&These errors are plotted as a function of the block size.
105 \&Also an analytical block average curve is plotted, assuming
106 \&that the autocorrelation is a sum of two exponentials.
107 \&The analytical curve for the block average is:
109 \&f(t) = sigma\fB *\fRsqrt(2/T ( alpha (tau_1 ((exp(\-t/tau_1) \- 1) tau_1/t + 1)) +
111 \& (1\-alpha) (tau_2 ((exp(\-t/tau_2) \- 1) tau_2/t + 1)))),
112 where T is the total time.
113 \&alpha, tau_1 and tau_2 are obtained by fitting f2(t) to error2.
114 \&When the actual block average is very close to the analytical curve,
115 \&the error is sigma\fB *\fRsqrt(2/T (a tau_1 + (1\-a) tau_2)).
116 \&The complete derivation is given in
117 \&B. Hess, J. Chem. Phys. 116:209\-217, 2002.
120 \&Option \fB \-bal\fR finds and subtracts the ultrafast "ballistic"
121 \&component from a hydrogen bond autocorrelation function by the fitting
122 \&of a sum of exponentials, as described in e.g.
123 \&O. Markovitch, J. Chem. Phys. 129:084505, 2008. The fastest term
124 \&is the one with the most negative coefficient in the exponential,
125 \&or with \fB \-d\fR, the one with most negative time derivative at time 0.
126 \&\fB \-nbalexp\fR sets the number of exponentials to fit.
129 \&Option \fB \-gem\fR fits bimolecular rate constants ka and kb
130 \&(and optionally kD) to the hydrogen bond autocorrelation function
131 \&according to the reversible geminate recombination model. Removal of
132 \&the ballistic component first is strongly advised. The model is presented in
133 \&O. Markovitch, J. Chem. Phys. 129:084505, 2008.
136 \&Option \fB \-filter\fR prints the RMS high\-frequency fluctuation
137 \&of each set and over all sets with respect to a filtered average.
138 \&The filter is proportional to cos(pi t/len) where t goes from \-len/2
139 \&to len/2. len is supplied with the option \fB \-filter\fR.
140 \&This filter reduces oscillations with period len/2 and len by a factor
141 \&of 0.79 and 0.33 respectively.
144 \&Option \fB \-g\fR fits the data to the function given with option
148 \&Option \fB \-power\fR fits the data to b ta, which is accomplished
149 \&by fitting to a t + b on log\-log scale. All points after the first
150 \&zero or with a negative value are ignored.
152 Option \fB \-luzar\fR performs a Luzar & Chandler kinetics analysis
153 \&on output from \fB g_hbond\fR. The input file can be taken directly
154 \&from \fB g_hbond \-ac\fR, and then the same result should be produced.
156 .BI "\-f" " graph.xvg"
160 .BI "\-ac" " autocorr.xvg"
164 .BI "\-msd" " msd.xvg"
168 .BI "\-cc" " coscont.xvg"
172 .BI "\-dist" " distr.xvg"
176 .BI "\-av" " average.xvg"
180 .BI "\-ee" " errest.xvg"
184 .BI "\-bal" " ballisitc.xvg"
188 .BI "\-g" " fitlog.log"
194 Print help info and quit
196 .BI "\-[no]version" "no "
197 Print version info and quit
199 .BI "\-nice" " int" " 0"
203 View output \fB .xvg\fR, \fB .xpm\fR, \fB .eps\fR and \fB .pdb\fR files
205 .BI "\-xvg" " enum" " xmgrace"
206 xvg plot formatting: \fB xmgrace\fR, \fB xmgr\fR or \fB none\fR
208 .BI "\-[no]time" "yes "
209 Expect a time in the input
211 .BI "\-b" " real" " \-1 "
212 First time to read from set
214 .BI "\-e" " real" " \-1 "
215 Last time to read from set
217 .BI "\-n" " int" " 1"
218 Read this number of sets separated by &
223 .BI "\-bw" " real" " 0.1 "
224 Binwidth for the distribution
226 .BI "\-errbar" " enum" " none"
227 Error bars for \fB \-av\fR: \fB none\fR, \fB stddev\fR, \fB error\fR or \fB 90\fR
229 .BI "\-[no]integrate" "no "
230 Integrate data function(s) numerically using trapezium rule
232 .BI "\-aver_start" " real" " 0 "
233 Start averaging the integral from here
235 .BI "\-[no]xydy" "no "
236 Interpret second data set as error in the y values for integrating
238 .BI "\-[no]regression" "no "
239 Perform a linear regression analysis on the data. If \fB \-xydy\fR is set a second set will be interpreted as the error bar in the Y value. Otherwise, if multiple data sets are present a multilinear regression will be performed yielding the constant A that minimize chi2 = (y \- A_0 x_0 \- A_1 x_1 \- ... \- A_N x_N)2 where now Y is the first data set in the input file and x_i the others. Do read the information at the option \fB \-time\fR.
241 .BI "\-[no]luzar" "no "
242 Do a Luzar and Chandler analysis on a correlation function and related as produced by \fB g_hbond\fR. When in addition the \fB \-xydy\fR flag is given the second and fourth column will be interpreted as errors in c(t) and n(t).
244 .BI "\-temp" " real" " 298.15"
245 Temperature for the Luzar hydrogen bonding kinetics analysis (K)
247 .BI "\-fitstart" " real" " 1 "
248 Time (ps) from which to start fitting the correlation functions in order to obtain the forward and backward rate constants for HB breaking and formation
250 .BI "\-fitend" " real" " 60 "
251 Time (ps) where to stop fitting the correlation functions in order to obtain the forward and backward rate constants for HB breaking and formation. Only with \fB \-gem\fR
253 .BI "\-smooth" " real" " \-1 "
254 If this value is = 0, the tail of the ACF will be smoothed by fitting it to an exponential function: y = A exp(\-x/tau)
256 .BI "\-filter" " real" " 0 "
257 Print the high\-frequency fluctuation after filtering with a cosine filter of this length
259 .BI "\-[no]power" "no "
262 .BI "\-[no]subav" "yes "
263 Subtract the average before autocorrelating
265 .BI "\-[no]oneacf" "no "
266 Calculate one ACF over all sets
268 .BI "\-acflen" " int" " \-1"
269 Length of the ACF, default is half the number of frames
271 .BI "\-[no]normalize" "yes "
274 .BI "\-P" " enum" " 0"
275 Order of Legendre polynomial for ACF (0 indicates none): \fB 0\fR, \fB 1\fR, \fB 2\fR or \fB 3\fR
277 .BI "\-fitfn" " enum" " none"
278 Fit function: \fB none\fR, \fB exp\fR, \fB aexp\fR, \fB exp_exp\fR, \fB vac\fR, \fB exp5\fR, \fB exp7\fR, \fB exp9\fR or \fB erffit\fR
280 .BI "\-ncskip" " int" " 0"
281 Skip this many points in the output file of correlation functions
283 .BI "\-beginfit" " real" " 0 "
284 Time where to begin the exponential fit of the correlation function
286 .BI "\-endfit" " real" " \-1 "
287 Time where to end the exponential fit of the correlation function, \-1 is until the end
292 More information about \fBGROMACS\fR is available at <\fIhttp://www.gromacs.org/\fR>.