-
Notifications
You must be signed in to change notification settings - Fork 4
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
17 changed files
with
743 additions
and
259 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,159 @@ | ||
| import numpy as np | ||
| import matplotlib.pyplot as mplt | ||
| from siriuspy.clientarch import PVDataSet, Time | ||
|
|
||
|
|
||
| def fit_pol(x, y, deg=1, polyfit=False): | ||
| # handle big timestamps | ||
| x = x - x[-1] | ||
| # considering y = b + ax | ||
| if polyfit: | ||
| (a, b), cov = np.polyfit(x, y, deg=deg, cov=True) | ||
| db_sqr, da_sqr = np.diag(cov) | ||
| else: | ||
| (b, a), details = np.polynomial.polynomial.polyfit(x, y, deg=deg, full=True) | ||
|
|
||
| residues = details[0] | ||
|
|
||
| lhs = np.polynomial.polynomial.polyvander(x, order-1) | ||
| scale = np.sqrt((lhs*lhs).sum(axis=0)) | ||
|
|
||
| lhs /= scale | ||
| lhs_inv = np.linalg.inv(np.dot(lhs.T, lhs)) | ||
| lhs_inv /= np.outer(scale, scale) | ||
|
|
||
| cov = lhs_inv * (residues/(lhs.shape[0] - order)) | ||
|
|
||
| da_sqr, db_sqr = np.diag(cov) | ||
|
|
||
| lt = - b / a | ||
| dlt_sqr_rel = da_sqr/a/a + db_sqr/b/b | ||
| dlt = np.sqrt(dlt_sqr_rel)*np.abs(lt) | ||
|
|
||
| return a, np.sqrt(da_sqr), b, np.sqrt(db_sqr), lt, dlt | ||
|
|
||
|
|
||
| # datasets | ||
| dataset = { | ||
| 0: { | ||
| 'title': '2023-01-23, 6-8h, N~4000', | ||
| 't_start': Time(2023, 1, 23, 6, 0, 0), | ||
| 't_stop': Time(2023, 1, 23, 8, 0, 0), | ||
| 'nrpts_fit': 4000, | ||
| 'lt_ylim': [0, 100], | ||
| 'err_ylim': [0, 1.0], | ||
| }, | ||
| 1: { | ||
| 'title': '2023-01-23, 10-11h', | ||
| 't_start': Time(2023, 1, 23, 10, 0, 0), | ||
| 't_stop': Time(2023, 1, 23, 11, 0, 0), | ||
| 'nrpts_fit': 200, | ||
| 'lt_ylim': [0, 100], | ||
| 'err_ylim': [0, 1.0], | ||
| }, | ||
| 2: { | ||
| 'title': '2023-01-23, 12h30-12h35', | ||
| 't_start': Time(2023, 1, 23, 12, 30, 0), | ||
| 't_stop': Time(2023, 1, 23, 12, 35, 0), | ||
| 'nrpts_fit': 200, | ||
| 'lt_ylim': [0, 200], | ||
| 'err_ylim': [-100.0, 100.0], | ||
| }, | ||
| 3: { | ||
| 'title': '2023-01-23, 18h35-19h25', | ||
| 't_start': Time(2023, 1, 23, 18, 35, 0), | ||
| 't_stop': Time(2023, 1, 23, 19, 25, 0), | ||
| 'nrpts_fit': 4000, | ||
| 'lt_ylim': [0, 100], | ||
| 'err_ylim': [0, 1.0], | ||
| }, | ||
| 4: { | ||
| 'title': '2023-02-08 8h15-20h', | ||
| 't_start': Time(2023, 2, 8, 8, 15, 0), | ||
| 't_stop': Time(2023, 2, 8, 20, 0, 0), | ||
| 'nrpts_fit': 500, | ||
| 'lt_ylim': [0, 100], | ||
| 'err_ylim': [0, 1.0], | ||
| }, | ||
| 5: { | ||
| 'title': '2023-02-14 16h-16h30', | ||
| 't_start': Time(2023, 2, 14, 16, 0, 0), | ||
| 't_stop': Time(2023, 2, 14, 16, 30, 0), | ||
| 'nrpts_fit': 1000, | ||
| 'lt_ylim': [0, 100], | ||
| 'err_ylim': [0, 1.0], | ||
| }, | ||
| } | ||
|
|
||
|
|
||
| pvnames = [ | ||
| 'SI-Glob:AP-CurrInfo:Lifetime-Mon', | ||
| 'SI-Glob:AP-CurrInfo:Current-Mon' | ||
| ] | ||
| pvds = PVDataSet(pvnames) | ||
| pvds.timeout = 30 | ||
|
|
||
| # for setid in dataset: | ||
| setid = 0 | ||
|
|
||
| # get archiver data | ||
| pvds.time_start = dataset[setid]['t_start'] | ||
| pvds.time_stop = dataset[setid]['t_stop'] | ||
| pvds.update() | ||
|
|
||
| ycurr = np.asarray(pvds['SI-Glob:AP-CurrInfo:Current-Mon'].value) | ||
| dcurr = np.hstack([0, np.diff(ycurr)]) | ||
| xcurr = np.asarray(pvds['SI-Glob:AP-CurrInfo:Current-Mon'].timestamp) | ||
| xcudt = [Time(t) for t in xcurr] | ||
| yltim = np.asarray(pvds['SI-Glob:AP-CurrInfo:Lifetime-Mon'].value) | ||
| xltim = np.asarray(pvds['SI-Glob:AP-CurrInfo:Lifetime-Mon'].timestamp) | ||
| xltdt = [Time(t) for t in xltim] | ||
| thres = 0.01 | ||
|
|
||
| # fit lifetime and errors | ||
| lenb = dataset[setid]['nrpts_fit'] # [#] | ||
| order = 2 | ||
| ltfit, lterr = [], [] | ||
| for i in range(lenb, len(ycurr)): | ||
| slc = slice(i-lenb, i) | ||
| idx = np.where(dcurr[slc] > thres)[0] | ||
| if idx.size: | ||
| slc = slice(idx[-1], i) | ||
| data = fit_pol(xcurr[slc], ycurr[slc], deg=order-1, polyfit=True) | ||
| ltfit.append(data[4]) | ||
| lterr.append(data[5]/data[4]) | ||
| ltfit, lterr = np.asarray(ltfit), np.asarray(lterr) | ||
|
|
||
| # plot | ||
| fig, ax = mplt.subplots() | ||
| fig.subplots_adjust(right=0.75) | ||
| tkw = dict(size=4, width=1.5) | ||
|
|
||
| ax.set_title(dataset[setid]['title'] + ', N~' + str(lenb)) | ||
| ax.set_xlabel('Time [s]') | ||
| ax.tick_params(axis='x', **tkw) | ||
| ax.xaxis.axis_date() | ||
|
|
||
| ax.set_ylabel('Current [mA]') | ||
| p, = ax.plot(xcudt, ycurr, 'b', label='Current') | ||
| ax.tick_params(axis='y', colors=p.get_color(), **tkw) | ||
| ax.yaxis.label.set_color(p.get_color()) | ||
|
|
||
| axlt = ax.twinx() | ||
| axlt.set_ylabel('Lifetime [h]') | ||
| po, = axlt.plot(xltdt, yltim/60/60, 'g', label='Lifetime Orig') | ||
| pf, = axlt.plot(xcudt[lenb:], ltfit/60/60, 'r', label='Lifetime Fit') | ||
| axlt.yaxis.label.set_color(po.get_color()) | ||
| axlt.tick_params(axis='y', colors=po.get_color(), **tkw) | ||
| axlt.set_ylim(dataset[setid]['lt_ylim']) | ||
|
|
||
| axerr = ax.twinx() | ||
| axerr.set_ylabel('Lifetime Fit Rel. Error') | ||
| axerr.spines['right'].set_position(("axes", 1.2)) | ||
| p, = axerr.plot(xcudt[lenb:], lterr, 'k', label='Lifetime Fit Error') | ||
| axerr.yaxis.label.set_color(p.get_color()) | ||
| axerr.tick_params(axis='y', colors=p.get_color(), **tkw) | ||
| axerr.set_ylim(dataset[setid]['err_ylim']) | ||
|
|
||
| fig.legend() | ||
| fig.show() |
Oops, something went wrong.