From quantum.analyst at gmail.com Tue Sep 15 21:34:31 2020 From: quantum.analyst at gmail.com (Elliott Sales de Andrade) Date: Tue, 15 Sep 2020 21:34:31 -0400 Subject: [Matplotlib-users] [ANN] Matplotlib 3.3.2 Message-ID: <93a668ef-21c6-3c2b-908e-67c69ef0a92a@gmail.com> Hi all, We are pleased to announce the release of 3.3.2. This is the second bug fix release of the 3.3.x series. Pre-built wheels are available for most major platforms, and can be installed using `pip install matplotlib==3.3.2`. Other packages may also be available already; please check with your preferred source. The 3.3.2 release represents the work of 14 authors over 24 pull requests, and we thank them for their contributions. This release contains several critical bug-fixes: * fix |Axis| scale on twinned |Axes| * fix auto-close of |Figure|s in nbagg * fix automatic title placement if |Axes| is off the |Figure| * fix autoranging of log scales with barstacked histogram * fix extra redraws when using |Button| or |TextBox| widgets * fix imshow with |LogNorm| and large vmin/vmax ranges * fix plotting Pandas |DataFrame| with string |MultiIndex| * fix |scatter| with |marker=''| * fix |scatter3d| color/linewidth re-projection * fix state of mode buttons in TkAgg backends * include license files in built distribution * reduce Visual C++ runtime requirements on Windows For further details, please see the What's new in Matplotlib 3.3.2 page: https://matplotlib.org/3.3.2/users/whats_new.html and the milestone on GitHub: https://github.com/matplotlib/matplotlib/milestone/57?closed=1 This release is signed by my GPG key. The fingerprint is: 23CA B59E 3332 F94D 26BE F037 8D86 E7FA E5EB 0C10 and it is also used to sign this message. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 833 bytes Desc: OpenPGP digital signature URL: From ABillings at axustech.com Mon Sep 21 13:10:46 2020 From: ABillings at axustech.com (Anthony Billings) Date: Mon, 21 Sep 2020 17:10:46 +0000 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... Message-ID: I wrote a program that gathers data and plots it (y) over time (x). I use Python 3.7.7, with matplotlib 3.2.2, on Windows 10 Pro. I have compiled the program using pyinstaller, into an .exe and put it on additional computers to run. The program has worked properly on all computers, except an HP i5, running Windows 10. I've attached a picture of both graphs to help explain my problem... Good and Bad Graph. The bad graph seems to display the data sporadically with long horizontal lines between data points. I have created a debug version that displays the data of the lists used for the x and y-axis and the lists are full with 80-110 values, so I know the data is being collected. It's just not displaying on the graph properly. I have tried uninstalling things and installing things, I have tweaked the priority in Task Manager, and I've tried running in SAFE mode to disable any extensions that may be interrupting in the background. Does anyone have ANY ideas or help? Thank you in advance! I have now tried installing Python 3.7.7 on the computer that displays the bad graph, and running the program directly from Python IDE and the display is still inaccurate. I'm at a total loss and would LOVE some help. Code snippet: def Laser_WaferScan(self, lst, display1, time_lst): """ Laser - WAFER SCAN PLOT """ waferScan = lst rpms = display1 # Displays the RPMs in plt.title waferScan_fullScan_time_lst = time_lst #### TESTING - Moved from main() 6.1.20 if gv.bandpass_yes == True: plt.subplot(3, 1, 1) else: plt.subplot(2, 1, 1) ########################################## # Dynamically setup graph axes using Plotting class waferScan_min_ylim, waferScan_max_ylim = self.Ylim(waferScan) plt.ylim(waferScan_min_ylim, waferScan_max_ylim) # Compare the last value added to the waferScan_fullScan_time_lst, and if it is greater than the current X-axis limit, make it the new X-axis limit if waferScan_fullScan_time_lst[-1] > gv.waferScan_max_xlim: gv.waferScan_max_xlim = waferScan_fullScan_time_lst[-1] plt.xlim(0, gv.waferScan_max_xlim) if gv.scan_name != None: # These conditional statements check to see which title should be used if gv.lowpass_yes == True: plt.title(f"{gv.scan_name}\nLowpass Filter Settings: N: {gv.lowpass_N} Wn: {gv.lowpass_Wn}\nRPMs = {rpms} Time to Scan = {round(gv.adjusted_time_to_scan,4)} sec", fontsize=12) else: plt.title(f"{gv.scan_name}\nRPMs = {rpms} Time to Scan = {round(gv.adjusted_time_to_scan,4)} sec", fontsize=12) else: if gv.lowpass_yes == True: plt.title(f"Lowpass Filter Settings: N: {gv.lowpass_N} Wn: {gv.lowpass_Wn}\nRPMs = {rpms} Time to Scan = {round(gv.adjusted_time_to_scan,4)} sec", fontsize=12) else: plt.title(f"RPMs = {rpms} Time to Scan = {round(gv.adjusted_time_to_scan,4)} sec", fontsize=12) plt.ylabel("Full Wafer Scan (volts)") plt.xlabel('Full Wafer Scan Time (milliseconds)', fontsize=10) plt.plot(waferScan_fullScan_time_lst, waferScan, label="Laser Readings", color='k') Email: ABillings at axustech.com 7001 West Erie Street Suite #1, Chandler, Arizona 85226 Phone: 480.705.8000 Direct Line: 480.240.5547 Website: www.axustech.com Surface Processing Technology for Semiconductors, III-V, and II-VI Materials, Advanced Packaging, TSV, MEMS, LED, Precision Optical, Photonics, Nanotechnology, and MicroMachining applications. We specialize in Legacy Equipment refurbishment, upgrading, modifications, repurposing, spare parts, and field support, and we OEM Gemini Slurry delivery systems, advanced wafer carriers, and robot-assisted polishers. We represent other OEMs including G&P Technologies and Trusval. We are the authorized sales, parts, service and support group for Gespac Controls and Moog Industrial Group products. Note: The information in this message and any attachments is confidential between the sender and the intended recipient. If you receive this message in error, please delete it immediately. Unauthorized use is strictly prohibited. -------------- next part -------------- An HTML attachment was scrubbed... URL: From silva at lma.cnrs-mrs.fr Mon Sep 21 15:06:50 2020 From: silva at lma.cnrs-mrs.fr (Fabrice Silva) Date: Mon, 21 Sep 2020 21:06:50 +0200 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: Le lundi 21 septembre 2020, Anthony Billings a ?crit : > I wrote a program that gathers data and plots it (y) over time (x). I > use Python 3.7.7, with matplotlib 3.2.2, on Windows 10 Pro. I have > compiled the program using pyinstaller, > into an .exe and put it on additional computers to run. The program > has worked properly on all computers, except an HP i5, running > Windows 10. I've attached a picture of both graphs to help explain my > problem... Good > and Bad Graph. > > > The bad graph seems to display the data sporadically with long > horizontal lines between data points. I have created a debug version > that displays the data of the lists used > for the x and y-axis and the lists are full with 80-110 values, so I > know the data is being collected. It's just not displaying on the > graph properly. I have tried uninstalling things and installing > things, I have tweaked the priority in Task Manager, and > I've tried running in SAFE mode to disable any extensions that may > be interrupting in the background. Does anyone have ANY ideas or > help? Thank you in advance! > > > I have now tried installing Python 3.7.7 on the computer that > displays the bad graph, and running the program directly from Python > IDE and the display is still inaccurate. > I'm at a total loss and would LOVE some help. Hi, Comparing the two plots, I am not sure matplotlib is the faulty one here. The y values just seem different between the two graphs. Did you tried to record the data and look at numerical values in order to check their consistency ? You may even transfer the data of the "bad plot" to the other computer (the good one), just to ensure whether it is a data problem or a plot problem... If they come from some experimental setup, maybe one of the devices is wrong... Fabrice -------------- next part -------------- An HTML attachment was scrubbed... URL: From ABillings at axustech.com Mon Sep 21 15:58:35 2020 From: ABillings at axustech.com (Anthony Billings) Date: Mon, 21 Sep 2020 19:58:35 +0000 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: The data is sampled from an external device based on the speed (rpms) of a spinning platen. The y-axis values are just set up by a function to adjust according to the range of the sampled data. As I mentioned in the OP the program works well on most computers. It?s just this HP, i5 10th gen I?m having trouble with. It may or may not be a matplotlib issue, but I don?t know what else it could be. I was referred to trying to change the epoch in matplotlib 3.3.2 and seeing if I could get it to work there. That did not fix the issue either. From: Matplotlib-users On Behalf Of Fabrice Silva Sent: Monday, September 21, 2020 12:07 PM To: matplotlib-users at python.org Subject: Re: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... Le lundi 21 septembre 2020, Anthony Billings a ?crit : I wrote a program that gathers data and plots it (y) over time (x). I use Python 3.7.7, with matplotlib 3.2.2, on Windows 10 Pro. I have compiled the program using pyinstaller, into an .exe and put it on additional computers to run. The program has worked properly on all computers, except an HP i5, running Windows 10. I've attached a picture of both graphs to help explain my problem... Good and Bad Graph. The bad graph seems to display the data sporadically with long horizontal lines between data points. I have created a debug version that displays the data of the lists used for the x and y-axis and the lists are full with 80-110 values, so I know the data is being collected. It's just not displaying on the graph properly. I have tried uninstalling things and installing things, I have tweaked the priority in Task Manager, and I've tried running in SAFE mode to disable any extensions that may be interrupting in the background. Does anyone have ANY ideas or help? Thank you in advance! I have now tried installing Python 3.7.7 on the computer that displays the bad graph, and running the program directly from Python IDE and the display is still inaccurate. I'm at a total loss and would LOVE some help. Hi, Comparing the two plots, I am not sure matplotlib is the faulty one here. The y values just seem different between the two graphs. Did you tried to record the data and look at numerical values in order to check their consistency ? You may even transfer the data of the "bad plot" to the other computer (the good one), just to ensure whether it is a data problem or a plot problem... If they come from some experimental setup, maybe one of the devices is wrong... Fabrice -------------- next part -------------- An HTML attachment was scrubbed... URL: From tcaswell at gmail.com Mon Sep 21 17:16:22 2020 From: tcaswell at gmail.com (Thomas Caswell) Date: Mon, 21 Sep 2020 17:16:22 -0400 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: Can you try plotting as plt.plot(waferScan_fullScan_time_lst, waferScan, '-o', label="Laser Readings", color='k') which will put a marker at each point to get a better sense of what the datapoints are and what parts are the interpretation is. Tom On Mon, Sep 21, 2020 at 3:58 PM Anthony Billings wrote: > The data is sampled from an external device based on the speed (rpms) of a > spinning platen. The y-axis values are just set up by a function to adjust > according to the range of the sampled data. > > As I mentioned in the OP the program works well on most computers. It?s > just this HP, i5 10th gen I?m having trouble with. It may or may not be a > matplotlib issue, but I don?t know what else it could be. > > I was referred to trying to change the epoch in matplotlib 3.3.2 and > seeing if I could get it to work there. That did not fix the issue either. > > > > > > *From:* Matplotlib-users axustech.com at python.org> *On Behalf Of *Fabrice Silva > *Sent:* Monday, September 21, 2020 12:07 PM > *To:* matplotlib-users at python.org > *Subject:* Re: [Matplotlib-users] Need assistance with inconsistent > matplotlib plots from computer to computer... > > > > Le lundi 21 septembre 2020, Anthony Billings a ?crit : > > I wrote a program that gathers data and plots it (y) over time (x). I use > Python 3.7.7, with matplotlib 3.2.2, on Windows 10 Pro. I have compiled the > program using pyinstaller, into an .exe and put it on additional computers > to run. The program has worked properly on all computers, except an HP i5, > running Windows 10. I've attached a picture of both graphs to help explain > my problem... Good and Bad Graph . > > The bad graph seems to display the data sporadically with long horizontal > lines between data points. I have created a debug version that displays the > data of the lists used for the x and y-axis and the lists are full with > 80-110 values, so I know the data is being collected. It's just not > displaying on the graph properly. I have tried uninstalling things and > installing things, I have tweaked the priority in Task Manager, and I've > tried running in SAFE mode to disable any extensions that may be > interrupting in the background. Does anyone have ANY ideas or help? Thank > you in advance! > > I have now tried installing Python 3.7.7 on the computer that displays the > bad graph, and running the program directly from Python IDE and the display > is still inaccurate. I'm at a total loss and would LOVE some help. > > > > Hi, > > Comparing the two plots, I am not sure matplotlib is the faulty one here. > > The y values just seem different between the two graphs. Did you tried to > record the data and look at numerical values in order to check their > consistency ? You may even transfer the data of the "bad plot" to the other > computer (the good one), just to ensure whether it is a data problem or a > plot problem... > > > > If they come from some experimental setup, maybe one of the devices is > wrong... > > > > Fabrice > _______________________________________________ > Matplotlib-users mailing list > Matplotlib-users at python.org > https://mail.python.org/mailman/listinfo/matplotlib-users > -- Thomas Caswell tcaswell at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From ABillings at axustech.com Mon Sep 21 18:06:55 2020 From: ABillings at axustech.com (Anthony Billings) Date: Mon, 21 Sep 2020 22:06:55 +0000 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: That was a neat idea! I added the ?-o? to the plot. My current workbench setup is currently tied to a power supply outputting 1.2v. The data is then read for approximately 0.108 (108 ms) - 0.1096 seconds (109.6 ms) and then displayed on the plot. I have a debug version running that shows there is a range of 160-170 entries in the list, all valid floats. Thus the plot that should be displayed is a straight line at 1.2 with ~160 dots. As the data was collected and displayed I saw variance in the plots. Most plots looked like this https://imgur.com/gallery/bEXc3t9, with about 7 or 8 equally distanced dots. This plot also shows an additional dot around the x=30 coordinate. Other plots had a ton of dots in between selected nodes, and then nothing between others, like this https://imgur.com/gallery/AezWLck. These 7/8 dots though are consistent throughout all of the plots. Which is curious? Why would only select values be displayed while others are not? Why on some scans are there LOTS of dots, and others barely any? -------------- next part -------------- An HTML attachment was scrubbed... URL: From ABillings at axustech.com Mon Sep 21 17:37:49 2020 From: ABillings at axustech.com (Anthony Billings) Date: Mon, 21 Sep 2020 21:37:49 +0000 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: That was a neat idea! I added the ?-o? to the plot. My current workbench setup is currently tied to a power supply outputting 1.2v. The data is then read for approximately 0.108 (108 ms) - 0.1096 seconds (109.6 ms) and then displayed on the plot. I have a debug version running that shows there is a range of 160-170 entries in the list, all valid floats. Thus the plot that should be displayed is a straight line at 1.2 with ~160 dots. As the data was collected and displayed I saw variance in the plots. Most plots looked like this https://imgur.com/gallery/bEXc3t9, with about 7 or 8 equally distanced dots. This plot also shows an additional dot around the x=30 coordinate. Other plots had a ton of dots in between selected nodes, and then nothing between others, like this https://imgur.com/gallery/AezWLck. These 7/8 dots though are consistent throughout all of the plots. Which is curious? Why would only select values be displayed while others are not? Why on some scans are there LOTS of dots, and others barely any? -------------- next part -------------- An HTML attachment was scrubbed... URL: From frank0734 at gmail.com Tue Sep 22 23:45:49 2020 From: frank0734 at gmail.com (Liang Wang) Date: Tue, 22 Sep 2020 23:45:49 -0400 Subject: [Matplotlib-users] symlog iffy when range max is below 1e0=1 Message-ID: Hi, I have a problem using 'symlog' when the range maximum is below 1e0=1. Perhaps I'm misusing it, possibly the linthresh option. Example notebook: https://gist.github.com/liangwang0734/fde9a033c3d8788075589e2f9fd87f0c Sample code: import numpy as np import matplotlib as mpl import matplotlib.pyplot as plt def test(start): linthresh = 10**(start+1) tmp = np.logspace(start, start + 3, 20) tmp = np.hstack((-tmp[::-1], tmp)) fig, ax = plt.subplots() ax.plot(tmp, marker='o') ax.set_yscale('symlog', linthresh=linthresh) ax1 = ax.twinx() ax1.plot(tmp, c='orange', ls='--') test(0) # OK, linear scale and symlog scale look different test(-3) # iffy, linear scale and symlog scale look the same [image: image.png] [image: image.png] -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image.png Type: image/png Size: 14506 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image.png Type: image/png Size: 11299 bytes Desc: not available URL: From frank0734 at gmail.com Wed Sep 23 00:48:36 2020 From: frank0734 at gmail.com (Liang Wang) Date: Wed, 23 Sep 2020 00:48:36 -0400 Subject: [Matplotlib-users] symlog iffy when range max is below 1e0=1 In-Reply-To: References: Message-ID: It turns out for yscale, I need to use linthreshy not linthresh. Problem solved. On Tue, Sep 22, 2020 at 11:45 PM Liang Wang wrote: > Hi, > > I have a problem using 'symlog' when the range maximum is below 1e0=1. > Perhaps I'm misusing it, possibly the linthresh option. > > Example notebook: > https://gist.github.com/liangwang0734/fde9a033c3d8788075589e2f9fd87f0c > > Sample code: > import numpy as np > import matplotlib as mpl > import matplotlib.pyplot as plt > > def test(start): > linthresh = 10**(start+1) > tmp = np.logspace(start, start + 3, 20) > tmp = np.hstack((-tmp[::-1], tmp)) > > fig, ax = plt.subplots() > ax.plot(tmp, marker='o') > ax.set_yscale('symlog', linthresh=linthresh) > ax1 = ax.twinx() > ax1.plot(tmp, c='orange', ls='--') > > test(0) # OK, linear scale and symlog scale look different > test(-3) # iffy, linear scale and symlog scale look the same > > [image: image.png] > > [image: image.png] > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image.png Type: image/png Size: 14506 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image.png Type: image/png Size: 11299 bytes Desc: not available URL: From ABillings at axustech.com Wed Sep 23 19:14:40 2020 From: ABillings at axustech.com (Anthony Billings) Date: Wed, 23 Sep 2020 23:14:40 +0000 Subject: [Matplotlib-users] Need assistance with inconsistent matplotlib plots from computer to computer... In-Reply-To: References: Message-ID: Problem Solved. McAfee Antivirus was causing the problem. -------------- next part -------------- An HTML attachment was scrubbed... URL: